Datasets:

---

MCG-NJU

/

OVBench

like 3

Follow

Multimedia Computing Group-Nanjing University 31

[ { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 921, "options": [ "A. [0.619, 0.415, 0.957, 0.991]", "B. [0.326, 0.678, 0.967, 0.875]", "C. [0.907, 0.476, 1.0, 0.904]", "D. [0.52, 0.121, 0.943, 0.974]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 965, "options": [ "A. answer phone -> ride (e.g., a bike, a car, a horse)", "B. shoot -> ride (e.g., a bike, a car, a horse)", "C. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group)", "D. hand shake -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 984, "options": [ "A. [0.047, 0.141, 0.656, 0.959]", "B. [0.0, 0.169, 0.431, 0.788]", "C. [0.245, 0.151, 0.827, 0.907]", "D. [0.0, 0.382, 0.806, 1.0]" ], "question": "Where is the person currently performing the drive (e.g., a car, a truck) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 991, "options": [ "A. drive (e.g., a car, a truck) -> talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)", "B. lift (a person) -> talk to (e.g., self, a person, a group) -> drive (e.g., a car, a truck)", "C. climb (e.g., a mountain) -> drive (e.g., a car, a truck) -> talk to (e.g., self, a person, a group)", "D. smoke -> talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1004, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.47, 0.104, 0.965, 1.0] location in the current frame performing the ride (e.g., a bike, a car, a horse)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1074, "options": [ "A. [0.226, 0.022, 0.847, 1.0]", "B. [0.449, 0.0, 0.858, 0.72]", "C. [0.444, 0.092, 1.0, 0.908]", "D. [0.095, 0.123, 0.661, 0.959]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1079, "options": [ "A. [0.0, 0.0, 1.0, 0.701]", "B. [0.207, 0.04, 0.857, 1.0]", "C. [0.383, 0.303, 0.874, 1.0]", "D. [0.551, 0.069, 1.0, 0.994]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1080, "options": [ "A. dress/put on clothing -> talk to (e.g., self, a person, a group)", "B. put down -> ride (e.g., a bike, a car, a horse)", "C. talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)", "D. talk to (e.g., self, a person, a group) -> get up" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1197, "options": [ "A. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group) -> drive (e.g., a car, a truck)", "B. talk to (e.g., self, a person, a group) -> drive (e.g., a car, a truck) -> carry/hold (an object)", "C. drive (e.g., a car, a truck) -> ride (e.g., a bike, a car, a horse) -> carry/hold (an object)", "D. drive (e.g., a car, a truck) -> eat -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1264, "options": [ "A. sit, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, ride (e.g., a bike, a car, a horse), listen to (a person)", "C. sit, watch (a person)", "D. sit, drive (e.g., a car, a truck)" ], "question": "What action is the person at the location [0.308, 0.191, 0.581, 0.754] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1485, "options": [ "A. [0.118, 0.165, 0.779, 0.972]", "B. [0.549, 0.105, 0.954, 0.978]", "C. [0.203, 0.378, 1.0, 0.733]", "D. [0.0, 0.383, 0.563, 0.689]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1508, "options": [ "A. [0.781, 0.115, 1.0, 0.778]", "B. [0.677, 0.341, 1.0, 1.0]", "C. [0.414, 0.031, 0.684, 0.95]", "D. [0.539, 0.097, 0.955, 0.981]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1550, "options": [ "A. [0.739, 0.0, 1.0, 0.714]", "B. [0.535, 0.197, 0.947, 0.992]", "C. [0.308, 0.427, 0.826, 1.0]", "D. [0.168, 0.064, 0.845, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1575, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.379, 0.168, 0.682, 0.98] location in the current frame performing the hug (a person)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1581, "options": [ "A. stand, grab (a person), talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. sit, talk to (e.g., self, a person, a group)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.111, 0.128, 0.51, 0.963] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1584, "options": [ "A. hug (a person)", "B. drive (e.g., a car, a truck)", "C. walk", "D. extract" ], "question": "What action is the person currently in the [0.135, 0.127, 0.526, 0.957] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1652, "options": [ "A. stand, watch (a person)", "B. stand, listen to (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, drive (e.g., a car, a truck)" ], "question": "What action is the person at the location [0.216, 0.053, 0.621, 0.963] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1703, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1768, "options": [ "A. walk -> bend/bow (at the waist)", "B. pull (an object) -> walk", "C. point to (an object) -> bend/bow (at the waist)", "D. walk -> kick (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1768 ]

[ { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 967, "options": [ "A. [0.032, 0.453, 0.176, 0.744]", "B. [0.723, 0.163, 0.988, 0.981]", "C. [0.506, 0.0, 1.0, 0.775]", "D. [0.479, 0.417, 1.0, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 994, "options": [ "A. play with kids", "B. walk and talk to (e.g., self, a person, a group)", "C. play musical instrument", "D. read" ], "question": "What action is the person currently in the [0.25, 0.277, 0.427, 0.992] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1003, "options": [ "A. [0.472, 0.626, 0.872, 1.0]", "B. [0.043, 0.0, 0.693, 1.0]", "C. [0.0, 0.28, 0.514, 0.933]", "D. [0.174, 0.359, 0.749, 0.986]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1008, "options": [ "A. [0.588, 0.321, 0.793, 0.985]", "B. [0.146, 0.066, 0.718, 0.67]", "C. [0.445, 0.014, 1.0, 0.59]", "D. [0.249, 0.224, 0.812, 0.782]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1038, "options": [ "A. [0.903, 0.047, 0.99, 0.951]", "B. [0.889, 0.611, 1.0, 1.0]", "C. [0.649, 0.205, 0.751, 0.61]", "D. [0.801, 0.332, 0.992, 0.982]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1040, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.092, 0.233, 0.382, 0.878] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1106, "options": [ "A. run/jog, carry/hold (an object), watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object), listen to (a person)", "D. stand, carry/hold (an object)" ], "question": "What action is the person at the location [0.474, 0.359, 0.569, 0.625] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1115, "options": [ "A. [0.311, 0.46, 0.339, 0.988]", "B. [0.918, 0.346, 0.997, 0.545]", "C. [0.233, 0.147, 0.206, 0.986]", "D. [0.028, 0.421, 0.207, 0.955]" ], "question": "What location in the frame is the person currently in the [0.362, 0.338, 0.578, 0.993] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1198, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1268, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "B. stand, carry/hold (an object), watch (a person)", "C. stand, carry/hold (an object)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.038, 0.144, 0.489, 0.639] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1452, "options": [ "A. [0.167, 0.291, 0.576, 0.992]", "B. [0.523, 0.148, 0.969, 0.952]", "C. [0.156, 0.271, 0.946, 1.0]", "D. [0.226, 0.01, 0.778, 0.955]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1471, "options": [ "A. [0.575, 0.304, 1.0, 0.991]", "B. [0.532, 0.533, 0.816, 1.0]", "C. [0.857, 0.151, 0.944, 0.82]", "D. [0.369, 0.16, 0.957, 0.911]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1484, "options": [ "A. [0.713, 0.269, 0.821, 0.656]", "B. [0.29, 0.021, 0.532, 0.995]", "C. [0.018, 0.0, 0.257, 1.0]", "D. [0.457, 0.0, 0.751, 0.748]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1564, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1692, "options": [ "A. stand, carry/hold (an object), listen to (a person)", "B. carry/hold (an object), fall down, watch (a person)", "C. stand, carry/hold (an object)", "D. walk, carry/hold (an object)" ], "question": "What action is the person at the location [0.366, 0.226, 0.502, 0.773] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1737, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.488, 0.362, 0.64, 0.795] location in the current frame performing the run/jog?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1737 ]

[ { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 968, "options": [ "A. put down [0.415, 0.328, 0.599, 1.000] -> carry/hold (an object) [0.739, 0.214, 0.715, 0.939] -> hand clap [0.512, 0.428, 0.584, 1.000]", "B. sing to (e.g., self, a person, a group) [0.501, 0.338, 0.785, 0.830] -> hand clap [0.521, 0.329, 0.816, 0.861] -> paint [0.593, 0.440, 0.945, 0.797]", "C. talk to (e.g., self, a person, a group) [0.695, 0.344, 0.945, 1.000] -> sing to (e.g., self, a person, a group) [0.415, 0.199, 0.812, 0.948] -> hand clap [0.422, 0.088, 0.592, 0.995]", "D. carry/hold (an object) [0.561, 0.257, 0.762, 0.976] -> hand clap [0.561, 0.257, 0.762, 0.976] -> sing to (e.g., self, a person, a group) [0.561, 0.257, 0.762, 0.976]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.561, 0.257, 0.762, 0.976] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1025, "options": [ "A. sit, listen to (a person), watch (a person)", "B. stand, walk, carry/hold (an object), watch (a person)", "C. sit, listen to (a person)", "D. walk, carry/hold (an object)" ], "question": "What action is the person at the location [0.13, 0.183, 0.299, 0.974] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1032, "options": [ "A. 3", "B. 4", "C. 5", "D. 6" ], "question": "How many people were performing the watch (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1169, "options": [ "A. [0.135, 0.0, 0.897, 0.968]", "B. [0.386, 0.0, 1.0, 0.891]", "C. [0.791, 0.393, 0.868, 0.839]", "D. [0.193, 0.281, 0.651, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1282, "options": [ "A. [0.313, 0.103, 0.711, 0.882]", "B. [0.471, 0.0, 0.925, 0.996]", "C. [0.339, 0.0, 0.49, 0.635]", "D. [0.111, 0.134, 0.644, 0.987]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1429, "options": [ "A. write [0.340, 0.245, 0.675, 1.000] -> talk to (e.g., self, a person, a group) [0.234, 0.337, 0.588, 1.000] -> carry/hold (an object) [0.374, 0.291, 0.633, 0.947]", "B. talk to (e.g., self, a person, a group) [0.396, 0.087, 0.965, 1.000] -> get up [0.091, 0.253, 0.683, 0.903] -> walk [0.142, 0.179, 0.783, 1.000]", "C. carry/hold (an object) [0.142, 0.054, 0.889, 1.000] -> talk to (e.g., self, a person, a group) [0.261, 0.177, 0.762, 1.000] -> jump/leap [0.422, 0.354, 0.728, 1.000]", "D. walk [0.269, 0.239, 0.77, 1.0] -> carry/hold (an object) [0.269, 0.239, 0.77, 1.0] -> talk to (e.g., self, a person, a group) [0.269, 0.239, 0.77, 1.0]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.269, 0.239, 0.77, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1436, "options": [ "A. stand, carry/hold (an object), hand clap, sing to (e.g., self, a person, a group), watch (a person)", "B. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "C. stand, listen to (a person), watch (a person)", "D. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.11, 0.0, 0.49, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1459, "options": [ "A. talk to (e.g., self, a person, a group) [0.215, 0.437, 0.405, 1.000] -> carry/hold (an object) [0.085, 0.459, 0.259, 1.000] -> put down [0.321, 0.619, 0.052, 0.852]", "B. catch (an object) [0.101, 0.304, 0.257, 0.920] -> talk to (e.g., self, a person, a group) [0.182, 0.464, 0.427, 0.999] -> carry/hold (an object) [0.031, 0.313, 0.369, 0.908]", "C. walk [0.000, 0.275, 0.072, 1.000] -> carry/hold (an object) [0.260, 0.504, 0.360, 0.929] -> brush teeth [0.170, 0.556, 0.534, 1.000]", "D. talk to (e.g., self, a person, a group) [0.265, 0.492, 0.424, 0.983] -> walk [0.125, 0.462, 0.229, 0.972] -> carry/hold (an object) [0.125, 0.462, 0.229, 0.972]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.125, 0.462, 0.229, 0.972] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1500, "options": [ "A. [0.075, 0.0, 0.438, 1.0]", "B. [0.341, 0.0, 0.671, 0.987]", "C. [0.554, 0.01, 0.845, 0.776]", "D. [0.621, 0.161, 0.884, 0.979]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1507, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1539, "options": [ "A. [0.102, 0.008, 0.755, 0.99]", "B. [0.566, 0.267, 0.945, 1.0]", "C. [0.146, 0.0, 0.665, 0.698]", "D. [0.277, 0.022, 0.883, 0.997]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1566, "options": [ "A. dig [0.250, 0.462, 0.382, 0.953] -> sing to (e.g., self, a person, a group) [0.569, 0.211, 0.426, 0.607] -> dance [0.516, 0.211, 0.748, 0.565]", "B. dance [0.465, 0.240, 0.629, 0.771] -> eat [0.601, 0.165, 0.634, 0.598] -> carry/hold (an object) [0.361, 0.356, 0.655, 0.741]", "C. carry/hold (an object) [0.447, 0.349, 0.577, 0.764] -> dance [0.447, 0.349, 0.577, 0.764] -> sing to (e.g., self, a person, a group) [0.447, 0.349, 0.577, 0.764]", "D. dance [0.492, 0.511, 0.548, 0.715] -> sail boat [0.334, 0.417, 0.609, 0.742] -> sing to (e.g., self, a person, a group) [0.464, 0.421, 0.424, 0.946]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.447, 0.349, 0.577, 0.764] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1611, "options": [ "A. [0.235, 0.299, 0.627, 1.0]", "B. [0.0, 0.0, 0.762, 0.882]", "C. [0.436, 0.003, 0.903, 0.992]", "D. [0.181, 0.056, 0.858, 0.983]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1618, "options": [ "A. carry/hold (an object) [0.418, 0.243, 0.625, 0.646] -> pull (an object) [0.530, 0.268, 0.544, 1.000] -> climb (e.g., a mountain) [0.601, 0.292, 0.618, 1.000]", "B. walk [0.458, 0.357, 0.568, 0.822] -> carry/hold (an object) [0.458, 0.357, 0.568, 0.822] -> pull (an object) [0.458, 0.357, 0.568, 0.822]", "C. carry/hold (an object) [0.640, 0.431, 0.376, 0.915] -> pull (an object) [0.453, 0.246, 0.463, 0.981] -> catch (an object) [0.342, 0.430, 0.524, 0.863]", "D. jump/leap [0.449, 0.426, 0.397, 0.827] -> walk [0.598, 0.240, 0.386, 0.823] -> pull (an object) [0.270, 0.395, 0.672, 0.913]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.458, 0.357, 0.568, 0.822] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1619, "options": [ "A. walk, carry/hold (an object)", "B. sit, listen to (a person)", "C. sit, listen to (a person), watch (a person)", "D. walk, listen to (a person)" ], "question": "What action is the person at the location [0.181, 0.437, 0.236, 0.685] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1682, "options": [ "A. 4", "B. 8", "C. 3", "D. 5" ], "question": "How many people were performing the carry/hold (an object) in the scene 300 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1700, "options": [ "A. close (e.g., a door, a box)", "B. climb (e.g., a mountain)", "C. take a photo", "D. carry/hold (an object) and crouch/kneel" ], "question": "What action is the person currently in the [0.693, 0.211, 0.75, 0.53] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1798, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1798 ]

[ { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1015, "options": [ "A. [0.663, 0.402, 0.739, 0.704]", "B. [0.777, 0.232, 0.708, 0.497]", "C. [0.407, 0.2, 0.698, 0.818]", "D. [0.947, 0.153, 0.998, 0.662]" ], "question": "Where was the person currently performing the bend/bow (at the waist) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1018, "options": [ "A. sit, ride (e.g., a bike, a car, a horse)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, talk to (e.g., self, a person, a group), watch (a person)", "D. sit, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.493, 0.304, 0.676, 0.892] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1071, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.189, 0.093, 0.947, 1.0] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1080, "options": [ "A. [0.189, 0.365, 0.652, 1.0]", "B. [0.417, 0.212, 1.0, 0.72]", "C. [0.045, 0.057, 0.738, 0.98]", "D. [0.197, 0.114, 0.937, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1092, "options": [ "A. exit -> carry/hold (an object)", "B. lift/pick up -> carry/hold (an object)", "C. carry/hold (an object) -> walk", "D. enter -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1176, "options": [ "A. stir", "B. carry/hold (an object) and cut", "C. bend/bow (at the waist) and carry/hold (an object)", "D. swim" ], "question": "What action is the person currently in the [0.251, 0.312, 0.656, 0.946] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1180, "options": [ "A. [0.239, 0.336, 0.62, 0.981]", "B. [0.025, 0.601, 0.504, 1.0]", "C. [0.53, 0.23, 0.594, 0.812]", "D. [0.073, 0.312, 0.471, 0.908]" ], "question": "Where was the person currently performing the cut in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1202, "options": [ "A. close (e.g., a door, a box) -> carry/hold (an object)", "B. drive (e.g., a car, a truck) -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. dress/put on clothing -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1227, "options": [ "A. open (e.g., a window, a car door) -> carry/hold (an object)", "B. talk to (e.g., self, a person, a group) -> lie/sleep", "C. turn (e.g., a screwdriver) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1248, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.062, 0.063, 0.91, 0.994] location in the current frame performing the eat?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1265, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> eat", "B. eat -> carry/hold (an object) -> drink", "C. read -> talk to (e.g., self, a person, a group) -> eat", "D. carry/hold (an object) -> eat -> press" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1271, "options": [ "A. shoot -> carry/hold (an object)", "B. eat -> hand wave", "C. eat -> throw", "D. carry/hold (an object) -> eat" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1415, "options": [ "A. crouch/kneel, watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. sit, watch (a person)", "D. sit, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.082, 0.137, 0.526, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1430, "options": [ "A. [0.451, 0.057, 0.941, 0.983]", "B. [0.566, 0.0, 0.968, 0.716]", "C. [0.534, 0.0, 0.663, 0.8]", "D. [0.096, 0.125, 0.504, 0.881]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1566, "options": [ "A. [0.49, 0.125, 0.852, 0.932]", "B. [0.128, 0.21, 0.506, 0.824]", "C. [0.226, 0.422, 0.356, 0.931]", "D. [0.336, 0.411, 0.774, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1604, "options": [ "A. sit, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "B. sit, watch (a person)", "C. sit, listen to (a person)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.126, 0.21, 0.52, 0.814] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1677, "options": [ "A. [0.048, 0.077, 0.866, 1.0]", "B. [0.0, 0.0, 1.0, 0.751]", "C. [0.536, 0.334, 1.0, 1.0]", "D. [0.261, 0.084, 0.986, 0.987]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1710, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1751, "options": [ "A. [0.381, 0.175, 0.675, 0.818]", "B. [0.674, 0.099, 0.811, 0.686]", "C. [0.442, 0.171, 0.964, 0.977]", "D. [0.236, 0.347, 0.783, 0.979]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1751 ]

[ { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 916, "options": [ "A. [0.815, 0.329, 0.959, 0.96]", "B. [0.621, 0.595, 0.816, 1.0]", "C. [1.0, 0.506, 1.0, 1.0]", "D. [0.018, 0.055, 0.703, 0.991]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 927, "options": [ "A. talk to (e.g., self, a person, a group) [0.386, 0.31, 0.558, 0.893] -> carry/hold (an object) [0.41, 0.287, 0.531, 0.898] -> touch (an object) [0.41, 0.287, 0.531, 0.898]", "B. hand wave [0.369, 0.135, 0.582, 0.928] -> carry/hold (an object) [0.303, 0.368, 0.714, 0.843] -> touch (an object) [0.549, 0.170, 0.496, 0.992]", "C. talk to (e.g., self, a person, a group) [0.458, 0.166, 0.453, 0.908] -> touch (an object) [0.346, 0.160, 0.549, 0.943] -> enter [0.310, 0.374, 0.687, 0.828]", "D. paint [0.519, 0.417, 0.701, 1.000] -> touch (an object) [0.237, 0.402, 0.582, 0.842] -> carry/hold (an object) [0.306, 0.222, 0.690, 0.760]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.41, 0.287, 0.531, 0.898] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1041, "options": [ "A. walk and carry/hold (an object)", "B. shoot", "C. enter", "D. ride (e.g., a bike, a car, a horse)" ], "question": "What action is the person currently in the [0.356, 0.352, 0.499, 0.698] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1042, "options": [ "A. [0.361, 0.279, 0.557, 0.682]", "B. [0.332, 0.253, 0.281, 0.542]", "C. [0.072, 0.368, 0.715, 0.949]", "D. [0.433, 0.192, 0.85, 0.978]" ], "question": "Where is the person currently performing the ride (e.g., a bike, a car, a horse) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1099, "options": [ "A. fishing [0.672, 0.296, 0.713, 1.000] -> carry/hold (an object) [0.682, 0.211, 0.752, 1.000] -> talk to (e.g., self, a person, a group) [0.341, 0.343, 0.681, 0.890]", "B. carry/hold (an object) [0.561, 0.322, 0.690, 1.000] -> talk to (e.g., self, a person, a group) [0.683, 0.393, 0.884, 0.911] -> swim [0.456, 0.196, 0.885, 0.846]", "C. walk [0.529, 0.344, 0.702, 0.994] -> carry/hold (an object) [0.529, 0.344, 0.702, 0.994] -> talk to (e.g., self, a person, a group) [0.529, 0.344, 0.702, 0.994]", "D. ride (e.g., a bike, a car, a horse) [0.545, 0.225, 0.708, 1.000] -> talk to (e.g., self, a person, a group) [0.720, 0.265, 0.559, 0.904] -> carry/hold (an object) [0.355, 0.266, 0.515, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.529, 0.344, 0.702, 0.994] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1122, "options": [ "A. sit, listen to (a person), watch (a person)", "B. stand, carry/hold (an object)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.019, 0.517, 0.498, 0.986] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1133, "options": [ "A. [0.313, 0.203, 0.663, 0.991]", "B. [0.48, 0.0, 0.794, 1.0]", "C. [0.24, 0.013, 0.519, 0.84]", "D. [0.069, 0.225, 0.479, 0.981]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1184, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.006, 0.1, 0.391, 0.998] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1218, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1252, "options": [ "A. [0.108, 0.439, 0.261, 0.942]", "B. [0.438, 0.004, 0.542, 1.0]", "C. [0.157, 0.175, 0.369, 0.924]", "D. [0.545, 0.065, 0.976, 0.975]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1260, "options": [ "A. sail boat -> carry/hold (an object)", "B. take (an object) from (a person) -> carry/hold (an object)", "C. touch (an object) -> carry/hold (an object)", "D. carry/hold (an object) -> answer phone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1281, "options": [ "A. stand, carry/hold (an object), pull (an object)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.08, 0.276, 0.317, 0.988] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1387, "options": [ "A. push (an object) -> carry/hold (an object) -> walk", "B. carry/hold (an object) -> martial art -> walk", "C. carry/hold (an object) -> dress/put on clothing -> walk", "D. walk -> carry/hold (an object) -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1442, "options": [ "A. crouch/kneel [0.343, 0.147, 0.744, 0.861] -> carry/hold (an object) [0.335, 0.159, 0.754, 0.847] -> cut [0.335, 0.159, 0.754, 0.847]", "B. cut [0.430, 0.257, 0.942, 1.000] -> dress/put on clothing [0.274, 0.071, 0.604, 0.826] -> carry/hold (an object) [0.428, 0.073, 0.869, 0.821]", "C. crouch/kneel [0.297, 0.310, 0.814, 0.953] -> ride (e.g., a bike, a car, a horse) [0.311, 0.142, 0.908, 1.000] -> carry/hold (an object) [0.172, 0.141, 0.634, 0.758]", "D. carry/hold (an object) [0.311, 0.166, 0.792, 0.764] -> text on/look at a cellphone [0.225, 0.000, 0.580, 0.718] -> cut [0.286, 0.276, 0.604, 0.787]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.335, 0.159, 0.754, 0.847] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1478, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.498, 0.244, 0.806, 0.995] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1535, "options": [ "A. [0.093, 0.28, 0.418, 0.994]", "B. [0.339, 0.602, 0.481, 0.86]", "C. [0.624, 0.517, 0.569, 0.917]", "D. [0.134, 0.848, 0.704, 1.0]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1540, "options": [ "A. catch (an object) -> carry/hold (an object)", "B. talk to (e.g., self, a person, a group) -> hand wave", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> kick (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1551, "options": [ "A. talk to (e.g., self, a person, a group) [0.321, 0.025, 0.543, 1.000] -> carry/hold (an object) [0.278, 0.014, 0.766, 0.793] -> write [0.287, 0.390, 0.810, 0.701]", "B. carry/hold (an object) [0.268, 0.19, 0.618, 0.873] -> crouch/kneel [0.268, 0.19, 0.618, 0.873] -> talk to (e.g., self, a person, a group) [0.268, 0.19, 0.618, 0.873]", "C. crouch/kneel [0.092, 0.046, 0.754, 0.685] -> carry/hold (an object) [0.371, 0.208, 0.559, 0.877] -> touch (an object) [0.103, 0.151, 0.670, 0.825]", "D. crouch/kneel [0.214, 0.157, 0.483, 0.709] -> paint [0.111, 0.338, 0.474, 0.729] -> carry/hold (an object) [0.180, 0.353, 0.486, 0.834]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.268, 0.19, 0.618, 0.873] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1553, "options": [ "A. carry/hold (an object) -> paint -> touch (an object) -> crouch/kneel", "B. touch (an object) -> crouch/kneel -> read -> carry/hold (an object)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> crouch/kneel -> touch (an object)", "D. crouch/kneel -> talk to (e.g., self, a person, a group) -> fishing -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1576, "options": [ "A. [0.543, 0.301, 0.942, 0.996]", "B. [0.262, 0.207, 0.608, 0.9]", "C. [0.0, 0.0, 0.54, 1.0]", "D. [0.38, 0.257, 0.819, 0.885]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1576, "options": [ "A. text on/look at a cellphone -> touch (an object) -> carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. touch (an object) -> crouch/kneel -> talk to (e.g., self, a person, a group) -> watch (e.g., TV)", "C. touch (an object) -> swim -> crouch/kneel -> carry/hold (an object)", "D. carry/hold (an object) -> crouch/kneel -> touch (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1639, "options": [ "A. [0.059, 0.038, 0.791, 0.714]", "B. [0.232, 0.125, 0.734, 0.982]", "C. [0.02, 0.288, 0.654, 0.992]", "D. [0.317, 0.0, 0.948, 0.99]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1662, "options": [ "A. talk to (e.g., self, a person, a group) [0.750, 0.345, 0.689, 0.963] -> hand wave [0.600, 0.360, 0.948, 0.838] -> carry/hold (an object) [0.628, 0.210, 0.997, 0.793]", "B. talk to (e.g., self, a person, a group) [0.617, 0.141, 1.000, 0.816] -> give/serve (an object) to (a person) [0.360, 0.116, 1.000, 1.000] -> hand clap [0.694, 0.027, 0.662, 0.835]", "C. carry/hold (an object) [0.559, 0.191, 0.86, 0.981] -> give/serve (an object) to (a person) [0.559, 0.191, 0.86, 0.981] -> talk to (e.g., self, a person, a group) [0.559, 0.191, 0.86, 0.981]", "D. carry/hold (an object) [0.586, 0.287, 0.923, 1.000] -> give/serve (an object) to (a person) [0.567, 0.218, 0.972, 0.802] -> give/serve (an object) to (a person) [0.537, 0.368, 0.733, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.559, 0.191, 0.86, 0.981] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1688, "options": [ "A. [0.526, 0.073, 0.977, 0.963]", "B. [0.505, 0.0, 0.951, 0.685]", "C. [0.813, 0.149, 1.0, 0.988]", "D. [0.393, 0.247, 0.545, 0.723]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1688 ]

[ { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 951, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.013, 0.127, 0.951, 0.985] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 954, "options": [ "A. read -> throw", "B. take a photo -> read", "C. carry/hold (an object) -> read", "D. crawl -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 977, "options": [ "A. read -> carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. open (e.g., a window, a car door) -> read -> talk to (e.g., self, a person, a group)", "C. grab (a person) -> read -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> read -> give/serve (an object) to (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1005, "options": [ "A. [0.499, 0.0, 0.889, 0.859]", "B. [0.0, 0.086, 0.405, 1.0]", "C. [0.431, 0.15, 0.535, 0.679]", "D. [0.312, 0.122, 0.817, 0.962]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1007, "options": [ "A. sit, listen to (a person)", "B. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "C. walk, listen to (a person), watch (a person)", "D. sit, watch (a person)" ], "question": "What action is the person at the location [0.01, 0.63, 0.362, 0.992] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1031, "options": [ "A. [0.647, 0.209, 1.0, 1.0]", "B. [0.611, 0.0, 0.608, 0.74]", "C. [0.454, 0.095, 0.905, 0.97]", "D. [0.001, 0.06, 0.512, 0.996]" ], "question": "Where was the person currently performing the sing to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1113, "options": [ "A. [0.473, 0.343, 1.0, 0.993]", "B. [0.247, 0.133, 0.904, 0.997]", "C. [0.796, 0.094, 0.885, 0.684]", "D. [0.554, 0.139, 0.873, 0.854]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1162, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1172, "options": [ "A. bend/bow (at the waist) [0.008, 0.091, 0.589, 0.993] -> grab (a person) [0.008, 0.091, 0.589, 0.993] -> talk to (e.g., self, a person, a group) [0.008, 0.091, 0.589, 0.993]", "B. lift/pick up [0.000, 0.051, 0.760, 1.000] -> bend/bow (at the waist) [0.000, 0.000, 0.592, 1.000] -> talk to (e.g., self, a person, a group) [0.156, 0.000, 0.453, 0.915]", "C. bend/bow (at the waist) [0.000, 0.029, 0.447, 1.000] -> grab (a person) [0.160, 0.069, 0.417, 1.000] -> stir [0.075, 0.024, 0.692, 1.000]", "D. bend/bow (at the waist) [0.000, 0.128, 0.701, 0.901] -> talk to (e.g., self, a person, a group) [0.171, 0.275, 0.689, 0.996] -> take a photo [0.000, 0.192, 0.729, 0.967]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.008, 0.091, 0.589, 0.993] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1220, "options": [ "A. talk to (e.g., self, a person, a group) [0.000, 0.057, 0.419, 1.000] -> carry/hold (an object) [0.136, 0.031, 0.374, 0.742] -> carry/hold (an object) [0.333, 0.094, 0.492, 0.631]", "B. talk to (e.g., self, a person, a group) [0.001, 0.06, 0.512, 0.996] -> bend/bow (at the waist) [0.203, 0.068, 0.501, 0.733] -> carry/hold (an object) [0.203, 0.068, 0.501, 0.733]", "C. talk to (e.g., self, a person, a group) [0.000, 0.000, 0.512, 0.895] -> bend/bow (at the waist) [0.108, 0.000, 0.594, 0.841] -> touch (an object) [0.056, 0.217, 0.390, 0.740]", "D. bend/bow (at the waist) [0.045, 0.097, 0.649, 0.877] -> talk to (e.g., self, a person, a group) [0.130, 0.247, 0.556, 1.000] -> take (an object) from (a person) [0.321, 0.000, 0.460, 0.787]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.203, 0.068, 0.501, 0.733] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1224, "options": [ "A. bend/bow (at the waist) [0.139, 0.000, 0.357, 1.000] -> paint [0.000, 0.000, 0.528, 1.000] -> talk to (e.g., self, a person, a group) [0.176, 0.165, 0.472, 0.899]", "B. bend/bow (at the waist) [0.0, 0.064, 0.526, 0.996] -> carry/hold (an object) [0.0, 0.064, 0.526, 0.996] -> talk to (e.g., self, a person, a group) [0.0, 0.064, 0.526, 0.996]", "C. enter [0.000, 0.000, 0.519, 0.853] -> bend/bow (at the waist) [0.062, 0.220, 0.562, 0.839] -> carry/hold (an object) [0.000, 0.094, 0.536, 1.000]", "D. carry/hold (an object) [0.082, 0.000, 0.465, 1.000] -> bend/bow (at the waist) [0.134, 0.252, 0.601, 1.000] -> bend/bow (at the waist) [0.189, 0.043, 0.664, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.064, 0.526, 0.996] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1226, "options": [ "A. bend/bow (at the waist) [0.000, 0.066, 0.412, 0.834] -> give/serve (an object) to (a person) [0.000, 0.006, 0.418, 1.000] -> carry/hold (an object) [0.172, 0.000, 0.462, 0.835] -> enter [0.144, 0.008, 0.562, 0.950]", "B. kick (an object) [0.000, 0.237, 0.635, 0.865] -> bend/bow (at the waist) [0.079, 0.216, 0.559, 0.879] -> give/serve (an object) to (a person) [0.043, 0.132, 0.496, 1.000] -> carry/hold (an object) [0.000, 0.076, 0.623, 1.000]", "C. talk to (e.g., self, a person, a group) [0.158, 0.198, 0.639, 1.000] -> open (e.g., a window, a car door) [0.000, 0.000, 0.351, 1.000] -> carry/hold (an object) [0.000, 0.081, 0.507, 1.000] -> give/serve (an object) to (a person) [0.101, 0.033, 0.428, 1.000]", "D. talk to (e.g., self, a person, a group) [0.004, 0.06, 0.544, 0.989] -> bend/bow (at the waist) [0.0, 0.06, 0.544, 0.996] -> carry/hold (an object) [0.0, 0.06, 0.544, 0.996] -> give/serve (an object) to (a person) [0.0, 0.06, 0.544, 0.996]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.06, 0.544, 0.996] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1298, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> hand wave", "C. talk to (e.g., self, a person, a group) -> crouch/kneel", "D. talk to (e.g., self, a person, a group) -> exit" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1305, "options": [ "A. sit, listen to (a person)", "B. walk, listen to (a person)", "C. sit, watch (a person)", "D. stand, carry/hold (an object), watch (a person)" ], "question": "What action is the person at the location [0.205, 0.283, 0.366, 0.92] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1380, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. hand wave -> carry/hold (an object)", "C. carry/hold (an object) -> text on/look at a cellphone", "D. kiss (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1440, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1504, "options": [ "A. talk to (e.g., self, a person, a group) [0.405, 0.415, 0.860, 0.980] -> carry/hold (an object) [0.242, 0.115, 0.834, 0.794] -> throw [0.126, 0.115, 0.904, 1.000]", "B. talk to (e.g., self, a person, a group) [0.183, 0.252, 0.676, 0.893] -> carry/hold (an object) [0.408, 0.331, 0.858, 0.975] -> fishing [0.103, 0.101, 0.710, 1.000]", "C. walk [0.25, 0.228, 0.726, 0.979] -> carry/hold (an object) [0.25, 0.228, 0.726, 0.979] -> talk to (e.g., self, a person, a group) [0.25, 0.228, 0.726, 0.979]", "D. shoot [0.110, 0.139, 0.573, 0.892] -> talk to (e.g., self, a person, a group) [0.115, 0.332, 0.537, 1.000] -> carry/hold (an object) [0.136, 0.057, 0.528, 0.912]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.25, 0.228, 0.726, 0.979] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1508, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "B. carry/hold (an object), lie/sleep", "C. sit, listen to (a person), watch (a person)", "D. sit, listen to (a person)" ], "question": "What action is the person at the location [0.214, 0.149, 0.659, 0.989] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1545, "options": [ "A. [0.231, 0.0, 1.0, 0.772]", "B. [0.498, 0.073, 1.0, 0.982]", "C. [0.749, 0.0, 0.98, 0.726]", "D. [0.337, 0.132, 0.844, 0.985]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1594, "options": [ "A. [0.329, 0.094, 0.923, 0.975]", "B. [0.934, 0.443, 1.0, 0.925]", "C. [0.915, 0.047, 0.945, 0.818]", "D. [0.638, 0.171, 0.997, 0.973]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1623, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.51, 0.316, 0.979, 0.985] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1651, "options": [ "A. carry/hold (an object), talk to (e.g., self, a person, a group), lie/sleep", "B. stand, watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, listen to (a person)" ], "question": "What action is the person at the location [0.0, 0.467, 0.667, 0.996] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1666, "options": [ "A. talk to (e.g., self, a person, a group)", "B. clink glass", "C. extract", "D. carry/hold (an object) and lie/sleep" ], "question": "What action is the person currently in the [0.179, 0.232, 0.915, 0.988] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1683, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1739, "options": [ "A. [0.493, 0.167, 0.89, 0.985]", "B. [0.008, 0.211, 0.433, 0.986]", "C. [0.402, 0.442, 0.807, 1.0]", "D. [0.334, 0.077, 1.0, 0.744]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1779, "options": [ "A. walk -> hand clap", "B. walk -> put down", "C. walk -> talk to (e.g., self, a person, a group)", "D. walk -> text on/look at a cellphone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1779 ]

[ { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 920, "options": [ "A. 3", "B. 2", "C. 7", "D. 4" ], "question": "How many people in the current frame are performing the action: dance?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 927, "options": [ "A. Yes", "B. No" ], "question": "Are there still 5 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1009, "options": [ "A. 6", "B. 3", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: dance?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1023, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.262, 0.279, 0.445, 0.874] location in the current frame performing the dance?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1034, "options": [ "A. 3", "B. 4", "C. 2", "D. 1" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1098, "options": [ "A. talk to (e.g., self, a person, a group) [0.434, 0.324, 0.539, 0.839] -> walk [0.464, 0.302, 0.559, 0.847] -> grab (a person) [0.464, 0.302, 0.559, 0.847]", "B. work on a computer [0.542, 0.284, 0.405, 0.954] -> grab (a person) [0.609, 0.111, 0.734, 0.808] -> talk to (e.g., self, a person, a group) [0.333, 0.394, 0.380, 0.838]", "C. press [0.339, 0.462, 0.613, 0.899] -> grab (a person) [0.465, 0.292, 0.748, 1.000] -> talk to (e.g., self, a person, a group) [0.472, 0.411, 0.631, 0.647]", "D. grab (a person) [0.471, 0.461, 0.642, 0.647] -> walk [0.663, 0.322, 0.511, 0.683] -> open (e.g., a window, a car door) [0.484, 0.279, 0.596, 0.863]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.464, 0.302, 0.559, 0.847] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1101, "options": [ "A. 4", "B. 6", "C. 7", "D. 3" ], "question": "How many people were performing the dance in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1103, "options": [ "A. carry/hold (an object) [0.471, 0.316, 0.506, 0.872] -> talk to (e.g., self, a person, a group) [0.261, 0.324, 0.689, 0.751] -> kiss (a person) [0.512, 0.342, 0.665, 0.854]", "B. extract [0.513, 0.408, 0.602, 0.678] -> carry/hold (an object) [0.358, 0.430, 0.561, 0.924] -> talk to (e.g., self, a person, a group) [0.614, 0.184, 0.661, 0.888]", "C. walk [0.437, 0.306, 0.525, 0.874] -> carry/hold (an object) [0.437, 0.306, 0.525, 0.874] -> talk to (e.g., self, a person, a group) [0.437, 0.306, 0.525, 0.874]", "D. hand shake [0.519, 0.309, 0.382, 0.688] -> carry/hold (an object) [0.337, 0.496, 0.672, 0.792] -> walk [0.300, 0.150, 0.411, 0.678]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.437, 0.306, 0.525, 0.874] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1123, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.258, 0.28, 0.843] location in the current frame performing the eat?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1129, "options": [ "A. [0.177, 0.311, 0.555, 0.885]", "B. [0.109, 0.093, 0.85, 0.875]", "C. [0.0, 0.076, 0.694, 0.686]", "D. [0.105, 0.0, 0.884, 0.589]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1242, "options": [ "A. [0.0, 0.0, 0.632, 0.781]", "B. [0.0, 0.493, 0.689, 1.0]", "C. [0.004, 0.326, 0.169, 0.883]", "D. [0.014, 0.212, 0.548, 0.887]" ], "question": "What location in the frame is the person currently in the [0.352, 0.128, 0.828, 0.92] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1319, "options": [ "A. 7", "B. 4", "C. 6", "D. 3" ], "question": "How many people were performing the dance in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1344, "options": [ "A. fight/hit (a person) -> hand clap", "B. touch (an object) -> fight/hit (a person)", "C. martial art -> push (an object)", "D. fight/hit (a person) -> martial art" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1378, "options": [ "A. lift (a person) [0.223, 0.241, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.046, 0.088, 0.724, 1.000] -> bend/bow (at the waist) [0.110, 0.000, 0.823, 0.883]", "B. talk to (e.g., self, a person, a group) [0.074, 0.060, 0.744, 1.000] -> fight/hit (a person) [0.238, 0.015, 1.000, 0.898] -> kick (a person) [0.341, 0.164, 0.892, 1.000]", "C. watch (e.g., TV) [0.289, 0.088, 0.850, 1.000] -> talk to (e.g., self, a person, a group) [0.360, 0.002, 0.839, 0.819] -> fight/hit (a person) [0.125, 0.000, 0.828, 1.000]", "D. bend/bow (at the waist) [0.211, 0.095, 0.914, 0.931] -> fight/hit (a person) [0.211, 0.095, 0.914, 0.931] -> talk to (e.g., self, a person, a group) [0.211, 0.095, 0.914, 0.931]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.211, 0.095, 0.914, 0.931] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1379, "options": [ "A. bend/bow (at the waist) -> talk to (e.g., self, a person, a group) -> take a photo", "B. talk to (e.g., self, a person, a group) -> bend/bow (at the waist) -> walk", "C. fight/hit (a person) -> eat -> talk to (e.g., self, a person, a group)", "D. bend/bow (at the waist) -> fight/hit (a person) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1392, "options": [ "A. take a photo -> bend/bow (at the waist)", "B. chop -> bend/bow (at the waist)", "C. read -> bend/bow (at the waist)", "D. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1406, "options": [ "A. dance, sing to (e.g., self, a person, a group)", "B. stand, talk to (e.g., self, a person, a group)", "C. sit, listen to (a person), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.063, 0.417, 0.306, 0.873] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1431, "options": [ "A. 4", "B. 7", "C. 6", "D. 8" ], "question": "How many people were performing the watch (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1525, "options": [ "A. carry/hold (an object)", "B. push (an object)", "C. fishing", "D. answer phone" ], "question": "What action is the person currently in the [0.353, 0.14, 0.84, 0.88] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1528, "options": [ "A. cook -> answer phone", "B. text on/look at a cellphone -> take a photo", "C. text on/look at a cellphone -> answer phone", "D. climb (e.g., a mountain) -> text on/look at a cellphone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1554, "options": [ "A. carry/hold (an object) -> walk", "B. exit -> walk", "C. lift/pick up -> walk", "D. carry/hold (an object) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1595, "options": [ "A. bend/bow (at the waist) [0.545, 0.267, 0.840, 0.844] -> talk to (e.g., self, a person, a group) [0.548, 0.370, 0.818, 0.808] -> lie/sleep [0.632, 0.370, 0.858, 0.969]", "B. paint [0.768, 0.352, 0.783, 0.781] -> touch (an object) [0.496, 0.325, 1.000, 0.889] -> bend/bow (at the waist) [0.560, 0.021, 1.000, 0.989]", "C. hand wave [0.462, 0.165, 1.000, 0.805] -> bend/bow (at the waist) [0.642, 0.404, 1.000, 0.820] -> talk to (e.g., self, a person, a group) [0.577, 0.042, 1.000, 0.800]", "D. bend/bow (at the waist) [0.609, 0.205, 0.942, 0.887] -> touch (an object) [0.609, 0.205, 0.942, 0.887] -> talk to (e.g., self, a person, a group) [0.609, 0.205, 0.942, 0.887]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.609, 0.205, 0.942, 0.887] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1600, "options": [ "A. talk to (e.g., self, a person, a group) [0.583, 0.162, 0.901, 0.888] -> bend/bow (at the waist) [0.567, 0.173, 0.902, 0.893] -> touch (an object) [0.567, 0.173, 0.902, 0.893]", "B. close (e.g., a door, a box) [0.661, 0.315, 0.952, 0.924] -> touch (an object) [0.662, 0.335, 0.877, 0.843] -> talk to (e.g., self, a person, a group) [0.737, 0.281, 0.949, 0.750]", "C. pull (an object) [0.515, 0.118, 0.879, 0.947] -> talk to (e.g., self, a person, a group) [0.769, 0.091, 0.718, 0.786] -> touch (an object) [0.594, 0.120, 0.981, 1.000]", "D. play musical instrument [0.713, 0.275, 0.916, 1.000] -> bend/bow (at the waist) [0.423, 0.236, 0.844, 1.000] -> talk to (e.g., self, a person, a group) [0.385, 0.236, 0.719, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.567, 0.173, 0.902, 0.893] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1693, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1705, "options": [ "A. [0.236, 0.463, 0.814, 1.0]", "B. [0.0, 0.351, 0.434, 0.639]", "C. [0.198, 0.166, 0.607, 0.867]", "D. [0.537, 0.411, 0.568, 0.477]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1765, "options": [ "A. [0.112, 0.433, 0.368, 0.861]", "B. [0.966, 0.0, 1.0, 1.0]", "C. [0.761, 0.125, 0.903, 0.878]", "D. [0.739, 0.332, 0.727, 0.962]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1765 ]

[ { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 916, "options": [ "A. [0.032, 0.311, 0.326, 1.0]", "B. [0.497, 0.07, 0.906, 0.981]", "C. [0.006, 0.102, 0.609, 0.98]", "D. [0.291, 0.298, 0.746, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 933, "options": [ "A. [0.057, 0.0, 1.0, 0.738]", "B. [0.322, 0.188, 0.967, 0.983]", "C. [0.147, 0.33, 0.673, 1.0]", "D. [0.048, 0.046, 0.812, 0.98]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1005, "options": [ "A. read [0.074, 0.000, 0.795, 1.000] -> work on a computer [0.063, 0.261, 0.695, 0.814] -> carry/hold (an object) [0.017, 0.015, 0.776, 1.000]", "B. carry/hold (an object) [0.066, 0.078, 0.8, 0.94] -> read [0.066, 0.078, 0.8, 0.94] -> talk to (e.g., self, a person, a group) [0.066, 0.078, 0.8, 0.94]", "C. answer phone [0.178, 0.033, 0.608, 0.928] -> read [0.240, 0.269, 0.739, 0.916] -> carry/hold (an object) [0.000, 0.091, 0.864, 1.000]", "D. talk to (e.g., self, a person, a group) [0.262, 0.230, 0.880, 1.000] -> read [0.110, 0.060, 0.684, 0.791] -> cook [0.083, 0.272, 0.838, 0.815]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.066, 0.078, 0.8, 0.94] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1027, "options": [ "A. talk to (e.g., self, a person, a group) [0.065, 0.086, 0.641, 0.898] -> carry/hold (an object) [0.201, 0.042, 0.546, 1.000] -> chop [0.117, 0.258, 0.738, 1.000]", "B. text on/look at a cellphone [0.405, 0.337, 0.501, 1.000] -> carry/hold (an object) [0.316, 0.353, 0.598, 0.963] -> talk to (e.g., self, a person, a group) [0.247, 0.068, 0.586, 0.944]", "C. talk to (e.g., self, a person, a group) [0.079, 0.070, 0.587, 1.000] -> catch (an object) [0.228, 0.326, 0.549, 0.968] -> carry/hold (an object) [0.209, 0.143, 0.770, 0.790]", "D. walk [0.264, 0.166, 0.68, 0.977] -> carry/hold (an object) [0.264, 0.166, 0.68, 0.977] -> talk to (e.g., self, a person, a group) [0.264, 0.166, 0.68, 0.977]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.264, 0.166, 0.68, 0.977] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1043, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.231, 0.115, 0.726, 0.961] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1092, "options": [ "A. close (e.g., a door, a box) [0.561, 0.261, 0.634, 0.892] -> sing to (e.g., self, a person, a group) [0.514, 0.311, 0.724, 0.937] -> walk [0.516, 0.403, 0.722, 0.726]", "B. walk [0.387, 0.215, 0.543, 0.877] -> carry/hold (an object) [0.387, 0.215, 0.543, 0.877] -> close (e.g., a door, a box) [0.387, 0.215, 0.543, 0.877]", "C. walk [0.265, 0.058, 0.485, 0.841] -> carry/hold (an object) [0.410, 0.074, 0.586, 0.873] -> close (e.g., a door, a box) [0.354, 0.164, 0.406, 1.000]", "D. carry/hold (an object) [0.409, 0.228, 0.647, 0.757] -> close (e.g., a door, a box) [0.572, 0.102, 0.536, 1.000] -> grab (a person) [0.245, 0.259, 0.595, 0.759]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.387, 0.215, 0.543, 0.877] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1099, "options": [ "A. open (e.g., a window, a car door) [0.493, 0.000, 0.601, 1.000] -> talk to (e.g., self, a person, a group) [0.201, 0.000, 0.803, 0.925] -> touch (an object) [0.231, 0.000, 0.522, 1.000]", "B. talk to (e.g., self, a person, a group) [0.337, 0.081, 0.715, 0.988] -> carry/hold (an object) [0.326, 0.116, 0.596, 0.97] -> touch (an object) [0.326, 0.116, 0.596, 0.97]", "C. touch (an object) [0.155, 0.030, 0.609, 1.000] -> sing to (e.g., self, a person, a group) [0.512, 0.239, 0.651, 0.883] -> carry/hold (an object) [0.257, 0.127, 0.759, 1.000]", "D. touch (an object) [0.398, 0.120, 0.434, 1.000] -> talk to (e.g., self, a person, a group) [0.441, 0.000, 0.834, 1.000] -> dig [0.143, 0.067, 0.611, 0.883]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.326, 0.116, 0.596, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1109, "options": [ "A. sit, listen to (a person)", "B. sit, listen to (a person), watch (a person)", "C. stand, talk to (e.g., self, a person, a group)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.033, 0.106, 0.316, 0.962] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1124, "options": [ "A. touch (an object) [0.421, 0.409, 0.591, 0.848] -> carry/hold (an object) [0.309, 0.406, 0.792, 0.717] -> open (e.g., a window, a car door) [0.389, 0.304, 0.869, 0.647]", "B. carry/hold (an object) [0.399, 0.409, 0.630, 0.703] -> push (an object) [0.602, 0.163, 0.724, 0.873] -> talk to (e.g., self, a person, a group) [0.479, 0.123, 0.815, 0.951]", "C. carry/hold (an object) [0.422, 0.27, 0.669, 0.813] -> touch (an object) [0.422, 0.27, 0.669, 0.813] -> talk to (e.g., self, a person, a group) [0.422, 0.27, 0.669, 0.813]", "D. talk to (e.g., self, a person, a group) [0.300, 0.290, 0.803, 0.618] -> carry/hold (an object) [0.544, 0.102, 0.682, 0.692] -> work on a computer [0.396, 0.183, 0.524, 0.721]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.422, 0.27, 0.669, 0.813] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1189, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1197, "options": [ "A. 4", "B. 7", "C. 5", "D. 3" ], "question": "How many people were performing the listen to (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1307, "options": [ "A. 7", "B. 4", "C. 3", "D. 6" ], "question": "How many people were performing the listen to (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1372, "options": [ "A. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "B. stand, talk to (e.g., self, a person, a group)", "C. stand, hand shake, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.166, 0.207, 0.307, 0.511] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1423, "options": [ "A. 6", "B. 3", "C. 4", "D. 5" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1441, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1573, "options": [ "A. sit, watch (e.g., TV), listen to (a person)", "B. sit, listen to (a person), watch (a person)", "C. sit, talk to (e.g., self, a person, a group)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.012, 0.54, 0.418, 0.949] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1595, "options": [ "A. watch (e.g., TV)", "B. read", "C. open (e.g., a window, a car door)", "D. dress/put on clothing" ], "question": "What action is the person currently in the [0.519, 0.6, 0.866, 0.972] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1604, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1705, "options": [ "A. stand, listen to (a person)", "B. sit, listen to (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.005, 0.508, 0.257, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1778, "options": [ "A. 3", "B. 4", "C. 5", "D. 6" ], "question": "How many people were performing the listen to (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1796, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.008, 0.025, 0.638, 0.981] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1796 ]

[ { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 959, "options": [ "A. [0.527, 0.208, 0.547, 0.657]", "B. [0.352, 0.472, 0.504, 0.718]", "C. [0.105, 0.104, 0.334, 0.98]", "D. [0.584, 0.314, 0.71, 0.9]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 999, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1002, "options": [ "A. [0.581, 0.181, 0.74, 0.684]", "B. [0.032, 0.0, 0.424, 0.923]", "C. [0.49, 0.342, 0.814, 1.0]", "D. [0.24, 0.093, 0.637, 0.996]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1104, "options": [ "A. 4", "B. 3", "C. 5", "D. 6" ], "question": "How many people were performing the dance in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1111, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. walk, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.108, 0.073, 0.533, 0.987] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1138, "options": [ "A. talk to (e.g., self, a person, a group) -> bend/bow (at the waist) -> catch (an object)", "B. drink -> play with pets -> bend/bow (at the waist)", "C. give/serve (an object) to (a person) -> bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "D. bend/bow (at the waist) -> talk to (e.g., self, a person, a group) -> play with pets" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1196, "options": [ "A. carry/hold (an object) -> smoke", "B. carry/hold (an object) -> walk", "C. push (another person) -> walk", "D. carry/hold (an object) -> brush teeth" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1264, "options": [ "A. push (an object)", "B. carry/hold (an object)", "C. smoke", "D. lie/sleep" ], "question": "What action is the person currently in the [0.021, 0.004, 0.941, 1.0] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1302, "options": [ "A. touch (an object) -> open (e.g., a window, a car door) -> talk to (e.g., self, a person, a group) -> walk", "B. walk -> give/serve (an object) to (a person) -> bend/bow (at the waist) -> touch (an object)", "C. walk -> bend/bow (at the waist) -> touch (an object) -> catch (an object)", "D. open (e.g., a window, a car door) -> bend/bow (at the waist) -> walk -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1335, "options": [ "A. dance, watch (a person)", "B. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "C. walk, listen to (a person), watch (a person)", "D. sit, touch (an object), watch (a person)" ], "question": "What action is the person at the location [0.031, 0.2, 0.546, 0.966] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1363, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)", "B. walk -> talk to (e.g., self, a person, a group) -> close (e.g., a door, a box)", "C. walk -> carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1391, "options": [ "A. [0.281, 0.077, 0.78, 0.989]", "B. [0.726, 0.394, 1.0, 1.0]", "C. [0.707, 0.116, 0.907, 0.676]", "D. [0.465, 0.152, 0.969, 0.975]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1416, "options": [ "A. talk to (e.g., self, a person, a group) -> kiss (a person)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)", "D. kick (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1454, "options": [ "A. [0.012, 0.021, 0.958, 0.984]", "B. [0.247, 0.363, 0.989, 1.0]", "C. [0.17, 0.0, 0.931, 0.715]", "D. [0.301, 0.077, 0.982, 0.968]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1557, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, talk to (e.g., self, a person, a group)", "C. stand, listen to (a person), watch (a person)", "D. stand, play musical instrument, watch (a person)" ], "question": "What action is the person at the location [0.024, 0.122, 0.624, 0.979] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1686, "options": [ "A. 5", "B. 3", "C. 7", "D. 4" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1774, "options": [ "A. [0.051, 0.093, 0.681, 0.975]", "B. [0.0, 0.0, 0.715, 0.73]", "C. [0.386, 0.095, 1.0, 1.0]", "D. [0.097, 0.054, 0.953, 1.0]" ], "question": "Where is the person currently performing the run/jog located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1774 ]

[ { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 914, "options": [ "A. carry/hold (an object)", "B. carry/hold (an object) and talk to (e.g., self, a person, a group)", "C. exit", "D. work on a computer" ], "question": "What action is the person currently in the [0.044, 0.127, 0.524, 0.892] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1013, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.121, 0.032, 0.568, 0.955] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1125, "options": [ "A. 2", "B. 5", "C. 4", "D. 3" ], "question": "How many people in the current frame are performing the action: run/jog?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1133, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1183, "options": [ "A. 3", "B. 5", "C. 4", "D. 6" ], "question": "How many people were performing the run/jog in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1192, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. walk, watch (a person)", "C. stand, listen to (a person), talk to (e.g., self, a person, a group)", "D. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.019, 0.095, 0.459, 0.982] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1224, "options": [ "A. carry/hold (an object) [0.820, 0.427, 0.711, 0.494] -> dance [0.663, 0.227, 0.882, 0.666] -> talk to (e.g., self, a person, a group) [0.868, 0.306, 0.861, 0.536]", "B. walk [0.751, 0.097, 0.847, 0.733] -> carry/hold (an object) [0.811, 0.419, 0.941, 0.382] -> swim [0.789, 0.140, 0.746, 0.520]", "C. walk [0.556, 0.219, 0.636, 0.657] -> carry/hold (an object) [0.556, 0.155, 0.833, 0.442] -> read [0.697, 0.097, 0.897, 0.396]", "D. carry/hold (an object) [0.736, 0.244, 0.823, 0.558] -> talk to (e.g., self, a person, a group) [0.736, 0.244, 0.823, 0.558] -> walk [0.726, 0.25, 0.816, 0.567]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.726, 0.25, 0.816, 0.567] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1285, "options": [ "A. press -> walk", "B. talk to (e.g., self, a person, a group) -> walk", "C. walk -> bend/bow (at the waist)", "D. talk to (e.g., self, a person, a group) -> exit" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1291, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1292, "options": [ "A. watch (e.g., TV) [0.154, 0.000, 0.802, 1.000] -> smoke [0.070, 0.000, 0.871, 0.883] -> talk to (e.g., self, a person, a group) [0.310, 0.140, 0.759, 0.727]", "B. talk to (e.g., self, a person, a group) [0.233, 0.269, 0.761, 0.819] -> smoke [0.212, 0.216, 0.760, 1.000] -> exit [0.047, 0.092, 0.537, 1.000]", "C. talk to (e.g., self, a person, a group) [0.180, 0.000, 0.719, 1.000] -> smoke [0.294, 0.298, 0.903, 0.989] -> exit [0.120, 0.203, 0.667, 1.000]", "D. smoke [0.186, 0.1, 0.706, 0.926] -> touch (an object) [0.186, 0.1, 0.706, 0.926] -> talk to (e.g., self, a person, a group) [0.186, 0.1, 0.706, 0.926]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.186, 0.1, 0.706, 0.926] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1302, "options": [ "A. 3", "B. 4", "C. 5", "D. 2" ], "question": "How many people in the current frame are performing the action: touch (an object)?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1312, "options": [ "A. walk [0.385, 0.158, 0.578, 0.756] -> carry/hold (an object) [0.385, 0.158, 0.578, 0.756] -> talk to (e.g., self, a person, a group) [0.385, 0.158, 0.578, 0.756]", "B. carry/hold (an object) [0.446, 0.337, 0.462, 0.595] -> carry/hold (an object) [0.506, 0.000, 0.633, 0.777] -> walk [0.218, 0.017, 0.738, 0.581]", "C. talk to (e.g., self, a person, a group) [0.243, 0.011, 0.633, 0.710] -> carry/hold (an object) [0.227, 0.279, 0.755, 0.778] -> hand clap [0.358, 0.000, 0.657, 0.809]", "D. carry/hold (an object) [0.304, 0.340, 0.652, 0.791] -> talk to (e.g., self, a person, a group) [0.234, 0.278, 0.757, 0.615] -> work on a computer [0.360, 0.151, 0.409, 0.849]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.385, 0.158, 0.578, 0.756] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1313, "options": [ "A. [0.894, 0.196, 0.72, 0.662]", "B. [0.725, 0.374, 0.975, 0.862]", "C. [0.023, 0.011, 0.321, 0.62]", "D. [0.814, 0.638, 1.0, 1.0]" ], "question": "What location in the frame is the person currently in the [0.189, 0.37, 0.546, 0.88] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1430, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> read", "B. exit -> talk to (e.g., self, a person, a group) -> read", "C. read -> hand clap -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> point to (an object) -> read" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1438, "options": [ "A. carry/hold (an object) [0.047, 0.126, 0.623, 0.939] -> open (e.g., a window, a car door) [0.047, 0.126, 0.623, 0.939] -> walk [0.138, 0.157, 0.393, 0.981]", "B. carry/hold (an object) [0.209, 0.048, 0.451, 0.920] -> open (e.g., a window, a car door) [0.075, 0.206, 0.478, 1.000] -> fall down [0.186, 0.258, 0.330, 1.000]", "C. carry/hold (an object) [0.000, 0.082, 0.728, 0.938] -> walk [0.294, 0.217, 0.280, 0.916] -> crouch/kneel [0.141, 0.020, 0.621, 0.977]", "D. touch (an object) [0.168, 0.106, 0.519, 1.000] -> walk [0.310, 0.152, 0.365, 0.994] -> open (e.g., a window, a car door) [0.000, 0.007, 0.652, 0.803]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.138, 0.157, 0.393, 0.981] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1467, "options": [ "A. 4", "B. 3", "C. 6", "D. 5" ], "question": "How many people were performing the listen to (a person) in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1477, "options": [ "A. [0.04, 0.022, 0.998, 0.983]", "B. [0.164, 0.102, 0.304, 0.52]", "C. [0.36, 0.005, 0.526, 0.308]", "D. [0.216, 0.308, 0.508, 0.806]" ], "question": "Where was the person currently performing the crouch/kneel in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1489, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1496, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> lift (a person)", "C. hit (an object) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> extract" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1533, "options": [ "A. [0.751, 0.172, 1.0, 1.0]", "B. [0.394, 0.372, 0.872, 1.0]", "C. [0.024, 0.05, 0.769, 1.0]", "D. [0.548, 0.187, 0.808, 0.853]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1597, "options": [ "A. 3", "B. 4", "C. 5", "D. 6" ], "question": "How many people were performing the touch (an object) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1635, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. touch (an object) -> talk to (e.g., self, a person, a group)", "C. lie/sleep -> carry/hold (an object)", "D. carry/hold (an object) -> play with pets" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1669, "options": [ "A. [0.0, 0.003, 0.711, 0.67]", "B. [0.0, 0.444, 0.655, 1.0]", "C. [0.089, 0.151, 0.657, 0.828]", "D. [0.526, 0.192, 0.716, 0.391]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1683, "options": [ "A. sit, talk to (e.g., self, a person, a group)", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, talk to (e.g., self, a person, a group)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.086, 0.156, 0.547, 0.983] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1714, "options": [ "A. open (e.g., a window, a car door) -> talk to (e.g., self, a person, a group)", "B. work on a computer -> talk to (e.g., self, a person, a group)", "C. lift/pick up -> walk", "D. walk -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1720, "options": [ "A. carry/hold (an object) [0.476, 0.143, 0.808, 0.971] -> talk to (e.g., self, a person, a group) [0.476, 0.143, 0.808, 0.971] -> walk [0.347, 0.125, 0.758, 0.969]", "B. walk [0.366, 0.313, 0.869, 1.000] -> talk to (e.g., self, a person, a group) [0.403, 0.234, 0.723, 1.000] -> drink [0.301, 0.115, 0.970, 0.891]", "C. close (e.g., a door, a box) [0.283, 0.220, 0.807, 0.990] -> walk [0.302, 0.000, 0.824, 1.000] -> talk to (e.g., self, a person, a group) [0.592, 0.269, 0.889, 1.000]", "D. talk to (e.g., self, a person, a group) [0.475, 0.022, 0.880, 0.986] -> shovel [0.355, 0.000, 0.808, 1.000] -> carry/hold (an object) [0.382, 0.228, 0.900, 0.908]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.347, 0.125, 0.758, 0.969] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1739, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1739 ]

[ { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1072, "options": [ "A. [0.174, 0.011, 0.986, 0.971]", "B. [0.136, 0.109, 0.496, 1.0]", "C. [0.432, 0.0, 1.0, 0.806]", "D. [0.165, 0.022, 0.794, 0.716]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1088, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.062, 0.128, 0.559, 0.992] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1091, "options": [ "A. carry/hold (an object) -> play musical instrument -> walk", "B. kiss (a person) -> carry/hold (an object) -> walk", "C. carry/hold (an object) -> walk -> push (an object)", "D. talk to (e.g., self, a person, a group) -> walk -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1094, "options": [ "A. shoot [0.230, 0.056, 0.891, 1.000] -> carry/hold (an object) [0.000, 0.000, 0.894, 0.808] -> talk to (e.g., self, a person, a group) [0.000, 0.212, 0.806, 1.000]", "B. walk [0.055, 0.079, 0.786, 0.985] -> carry/hold (an object) [0.055, 0.079, 0.786, 0.985] -> talk to (e.g., self, a person, a group) [0.055, 0.079, 0.786, 0.985]", "C. take a photo [0.250, 0.175, 0.894, 0.937] -> walk [0.005, 0.000, 0.648, 1.000] -> talk to (e.g., self, a person, a group) [0.058, 0.000, 0.917, 0.939]", "D. walk [0.000, 0.217, 0.684, 0.839] -> enter [0.178, 0.057, 0.799, 0.852] -> carry/hold (an object) [0.066, 0.099, 0.769, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.055, 0.079, 0.786, 0.985] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1106, "options": [ "A. [0.098, 0.0, 0.765, 0.719]", "B. [0.076, 0.318, 0.696, 0.863]", "C. [0.0, 0.005, 0.383, 0.984]", "D. [0.317, 0.069, 0.959, 0.983]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1130, "options": [ "A. chop -> walk -> carry/hold (an object)", "B. drink -> carry/hold (an object) -> walk", "C. walk -> carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. walk -> answer phone -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1135, "options": [ "A. walk -> talk to (e.g., self, a person, a group)", "B. hand shake -> walk", "C. play with kids -> walk", "D. work on a computer -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1177, "options": [ "A. eat", "B. carry/hold (an object)", "C. shovel", "D. press" ], "question": "What action is the person currently in the [0.64, 0.23, 0.991, 0.957] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1179, "options": [ "A. [0.526, 0.249, 0.867, 1.0]", "B. [0.364, 0.488, 0.323, 1.0]", "C. [0.189, 0.116, 0.648, 0.683]", "D. [0.372, 0.23, 0.515, 0.92]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1273, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1365, "options": [ "A. 6", "B. 3", "C. 5", "D. 4" ], "question": "How many people were performing the watch (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1420, "options": [ "A. carry/hold (an object), crouch/kneel, listen to (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.0, 0.155, 0.229, 0.658] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1463, "options": [ "A. play with kids -> talk to (e.g., self, a person, a group)", "B. eat -> walk", "C. talk to (e.g., self, a person, a group) -> walk", "D. talk to (e.g., self, a person, a group) -> take (an object) from (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1510, "options": [ "A. talk to (e.g., self, a person, a group) [0.123, 0.000, 0.829, 1.000] -> fight/hit (a person) [0.000, 0.048, 0.693, 1.000] -> carry/hold (an object) [0.041, 0.178, 0.918, 0.845]", "B. talk to (e.g., self, a person, a group) [0.203, 0.203, 0.577, 0.851] -> walk [0.192, 0.031, 0.949, 1.000] -> fight/hit (a person) [0.184, 0.000, 0.832, 0.996]", "C. walk [0.141, 0.081, 0.795, 1.000] -> crouch/kneel [0.254, 0.150, 0.642, 1.000] -> carry/hold (an object) [0.261, 0.005, 0.683, 0.987]", "D. walk [0.099, 0.159, 0.758, 0.98] -> carry/hold (an object) [0.099, 0.159, 0.758, 0.98] -> talk to (e.g., self, a person, a group) [0.099, 0.159, 0.758, 0.98]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.099, 0.159, 0.758, 0.98] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1526, "options": [ "A. [0.06, 0.001, 0.937, 0.985]", "B. [0.0, 0.278, 0.811, 1.0]", "C. [0.0, 0.31, 0.463, 1.0]", "D. [0.0, 0.033, 0.607, 1.0]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1527, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.007, 0.044, 0.62, 0.993] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1530, "options": [ "A. hit (an object) -> talk to (e.g., self, a person, a group)", "B. get up -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. paint -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1605, "options": [ "A. [0.59, 0.187, 0.984, 0.754]", "B. [0.449, 0.386, 0.992, 0.991]", "C. [0.003, 0.003, 0.847, 0.986]", "D. [0.307, 0.0, 1.0, 0.64]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1622, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "B. sit, carry/hold (an object), listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.003, 0.182, 0.483, 0.989] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1635, "options": [ "A. [0.0, 0.0, 0.308, 1.0]", "B. [0.3, 0.0, 1.0, 0.7]", "C. [0.159, 0.255, 0.719, 0.99]", "D. [0.21, 0.005, 0.996, 0.988]" ], "question": "Where was the person currently performing the get up in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1709, "options": [ "A. [0.511, 0.154, 0.652, 0.569]", "B. [0.066, 0.337, 0.389, 0.983]", "C. [0.177, 0.263, 0.591, 0.878]", "D. [0.159, 0.064, 0.382, 0.837]" ], "question": "What location in the frame is the person currently in the [0.445, 0.211, 0.715, 0.943] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1709, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.445, 0.211, 0.715, 0.943] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1712, "options": [ "A. talk to (e.g., self, a person, a group) [0.384, 0.227, 0.848, 1.000] -> grab (a person) [0.433, 0.306, 0.865, 1.000] -> walk [0.364, 0.343, 0.802, 0.977]", "B. extract [0.478, 0.172, 0.620, 0.901] -> carry/hold (an object) [0.581, 0.354, 0.889, 0.919] -> walk [0.289, 0.302, 0.781, 1.000]", "C. walk [0.441, 0.224, 0.753, 0.957] -> carry/hold (an object) [0.441, 0.224, 0.753, 0.957] -> talk to (e.g., self, a person, a group) [0.441, 0.224, 0.753, 0.957]", "D. carry/hold (an object) [0.444, 0.244, 0.784, 0.803] -> talk to (e.g., self, a person, a group) [0.260, 0.102, 0.718, 0.856] -> chop [0.603, 0.067, 0.730, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.441, 0.224, 0.753, 0.957] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1716, "options": [ "A. carry/hold (an object) [0.104, 0.107, 0.381, 0.923] -> talk to (e.g., self, a person, a group) [0.104, 0.107, 0.381, 0.923] -> walk [0.344, 0.138, 0.658, 0.818]", "B. talk to (e.g., self, a person, a group) [0.207, 0.000, 0.483, 0.881] -> answer phone [0.479, 0.031, 0.632, 0.628] -> carry/hold (an object) [0.023, 0.171, 0.565, 1.000]", "C. talk to (e.g., self, a person, a group) [0.056, 0.194, 0.321, 1.000] -> carry/hold (an object) [0.155, 0.092, 0.342, 1.000] -> chop [0.326, 0.063, 0.769, 0.749]", "D. carry/hold (an object) [0.095, 0.254, 0.529, 0.926] -> talk to (e.g., self, a person, a group) [0.000, 0.074, 0.455, 1.000] -> pull (an object) [0.256, 0.138, 0.496, 0.773]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.344, 0.138, 0.658, 0.818] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1722, "options": [ "A. carry/hold (an object) [0.441, 0.08, 1.0, 0.982] -> touch (an object) [0.441, 0.08, 1.0, 0.982] -> talk to (e.g., self, a person, a group) [0.441, 0.08, 1.0, 0.982]", "B. talk to (e.g., self, a person, a group) [0.598, 0.000, 0.832, 0.959] -> talk to (e.g., self, a person, a group) [0.426, 0.243, 1.000, 0.932] -> carry/hold (an object) [0.604, 0.148, 0.837, 0.899]", "C. talk to (e.g., self, a person, a group) [0.257, 0.000, 0.984, 0.836] -> kiss (a person) [0.423, 0.000, 0.981, 0.964] -> touch (an object) [0.588, 0.100, 1.000, 0.992]", "D. touch (an object) [0.638, 0.071, 1.000, 1.000] -> turn (e.g., a screwdriver) [0.472, 0.000, 0.898, 1.000] -> talk to (e.g., self, a person, a group) [0.513, 0.014, 0.941, 0.843]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.441, 0.08, 1.0, 0.982] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1763, "options": [ "A. 4", "B. 5", "C. 3", "D. 6" ], "question": "How many people were performing the stand in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 952, 1763 ]

[ { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 917, "options": [ "A. sit, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person)", "C. stand, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.123, 0.0, 0.465, 0.985] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 938, "options": [ "A. walk -> brush teeth", "B. walk -> martial art", "C. walk -> answer phone", "D. walk -> dig" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1156, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, listen to (a person)", "C. sit, listen to (a person)", "D. sit, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.0, 0.137, 0.325, 0.895] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1194, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1252, "options": [ "A. drink -> talk to (e.g., self, a person, a group) -> lie/sleep", "B. grab (a person) -> answer phone -> talk to (e.g., self, a person, a group)", "C. lie/sleep -> touch (an object) -> talk to (e.g., self, a person, a group)", "D. answer phone -> talk to (e.g., self, a person, a group) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1268, "options": [ "A. [0.067, 0.0, 0.958, 0.724]", "B. [0.028, 0.001, 0.53, 0.969]", "C. [0.283, 0.145, 0.996, 1.0]", "D. [0.154, 0.006, 0.707, 0.994]" ], "question": "Where was the person currently performing the answer phone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1271, "options": [ "A. answer phone -> lift (a person)", "B. answer phone -> talk to (e.g., self, a person, a group)", "C. answer phone -> paint", "D. talk to (e.g., self, a person, a group) -> climb (e.g., a mountain)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1295, "options": [ "A. lie/sleep -> lie/sleep", "B. grab (a person) -> answer phone", "C. answer phone -> lie/sleep", "D. kick (a person) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1325, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. walk, carry/hold (an object)", "C. stand, listen to (a person), watch (a person)", "D. sit, carry/hold (an object)" ], "question": "What action is the person at the location [0.142, 0.0, 0.473, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1375, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1384, "options": [ "A. walk -> touch (an object)", "B. walk -> carry/hold (an object)", "C. carry/hold (an object) -> hand shake", "D. walk -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1399, "options": [ "A. walk [0.419, 0.092, 0.429, 0.707] -> carry/hold (an object) [0.483, 0.034, 0.713, 0.724] -> cut [0.603, 0.404, 0.631, 0.954]", "B. walk [0.439, 0.217, 0.557, 0.888] -> carry/hold (an object) [0.439, 0.217, 0.557, 0.888] -> talk to (e.g., self, a person, a group) [0.439, 0.217, 0.557, 0.888]", "C. carry/hold (an object) [0.368, 0.115, 0.705, 0.961] -> ride (e.g., a bike, a car, a horse) [0.407, 0.414, 0.483, 0.747] -> walk [0.477, 0.168, 0.476, 0.875]", "D. walk [0.579, 0.197, 0.631, 0.915] -> carry/hold (an object) [0.442, 0.381, 0.474, 0.858] -> lift (a person) [0.611, 0.040, 0.584, 0.896]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.439, 0.217, 0.557, 0.888] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1406, "options": [ "A. talk to (e.g., self, a person, a group) [0.377, 0.147, 0.336, 0.773] -> close (e.g., a door, a box) [0.323, 0.000, 0.784, 0.922] -> carry/hold (an object) [0.223, 0.000, 0.461, 0.855]", "B. play board game [0.507, 0.148, 0.532, 0.777] -> carry/hold (an object) [0.219, 0.143, 0.428, 0.951] -> talk to (e.g., self, a person, a group) [0.404, 0.197, 0.310, 0.748]", "C. carry/hold (an object) [0.439, 0.071, 0.794, 0.795] -> run/jog [0.298, 0.102, 0.509, 0.685] -> walk [0.391, 0.000, 0.444, 0.705]", "D. talk to (e.g., self, a person, a group) [0.346, 0.015, 0.493, 0.744] -> walk [0.409, 0.015, 0.618, 0.862] -> carry/hold (an object) [0.409, 0.015, 0.618, 0.862]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.409, 0.015, 0.618, 0.862] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1409, "options": [ "A. [0.537, 0.083, 0.976, 0.99]", "B. [0.149, 0.14, 0.703, 0.921]", "C. [0.403, 0.042, 0.812, 0.882]", "D. [0.145, 0.0, 0.736, 0.699]" ], "question": "What location in the frame is the person currently in the [0.0, 0.295, 0.35, 1.0] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1422, "options": [ "A. talk to (e.g., self, a person, a group) [0.367, 0.068, 0.758, 0.899] -> carry/hold (an object) [0.39, 0.12, 0.795, 0.911] -> read [0.39, 0.12, 0.795, 0.911]", "B. pull (an object) [0.496, 0.000, 0.893, 0.878] -> read [0.482, 0.041, 0.711, 0.775] -> carry/hold (an object) [0.247, 0.184, 0.863, 0.928]", "C. carry/hold (an object) [0.571, 0.000, 0.820, 1.000] -> talk to (e.g., self, a person, a group) [0.352, 0.000, 0.708, 0.722] -> turn (e.g., a screwdriver) [0.430, 0.000, 0.822, 0.900]", "D. shovel [0.174, 0.082, 0.948, 0.946] -> carry/hold (an object) [0.432, 0.215, 0.628, 0.825] -> read [0.242, 0.248, 0.644, 0.956]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.39, 0.12, 0.795, 0.911] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1423, "options": [ "A. push (another person) [0.192, 0.160, 0.408, 0.472] -> talk to (e.g., self, a person, a group) [0.008, 0.394, 0.225, 0.659] -> bend/bow (at the waist) [0.283, 0.331, 0.206, 0.688]", "B. bend/bow (at the waist) [0.194, 0.264, 0.333, 0.557] -> carry/hold (an object) [0.194, 0.264, 0.333, 0.557] -> talk to (e.g., self, a person, a group) [0.194, 0.264, 0.333, 0.557]", "C. bend/bow (at the waist) [0.208, 0.109, 0.465, 0.739] -> talk to (e.g., self, a person, a group) [0.034, 0.464, 0.297, 0.470] -> open (e.g., a window, a car door) [0.266, 0.250, 0.459, 0.396]", "D. catch (an object) [0.048, 0.440, 0.209, 0.532] -> talk to (e.g., self, a person, a group) [0.063, 0.132, 0.447, 0.424] -> carry/hold (an object) [0.169, 0.230, 0.427, 0.584]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.194, 0.264, 0.333, 0.557] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1424, "options": [ "A. read [0.225, 0.063, 0.627, 1.000] -> talk to (e.g., self, a person, a group) [0.326, 0.101, 0.715, 0.948] -> sing to (e.g., self, a person, a group) [0.211, 0.000, 0.977, 0.754]", "B. play with kids [0.480, 0.045, 0.777, 0.900] -> read [0.565, 0.303, 0.804, 0.905] -> talk to (e.g., self, a person, a group) [0.292, 0.233, 0.842, 1.000]", "C. talk to (e.g., self, a person, a group) [0.367, 0.068, 0.758, 0.899] -> carry/hold (an object) [0.39, 0.12, 0.795, 0.911] -> read [0.39, 0.12, 0.795, 0.911]", "D. exit [0.329, 0.279, 0.742, 1.000] -> talk to (e.g., self, a person, a group) [0.354, 0.052, 0.594, 0.789] -> carry/hold (an object) [0.513, 0.087, 0.943, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.39, 0.12, 0.795, 0.911] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1448, "options": [ "A. answer phone", "B. work on a computer", "C. sail boat", "D. bend/bow (at the waist)" ], "question": "What action is the person currently in the [0.175, 0.252, 0.336, 0.58] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1450, "options": [ "A. bend/bow (at the waist), work on a computer", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. walk, carry/hold (an object)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.189, 0.268, 0.348, 0.57] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1566, "options": [ "A. [0.021, 0.141, 0.52, 0.996]", "B. [0.293, 0.0, 0.777, 0.782]", "C. [0.583, 0.011, 0.92, 0.986]", "D. [0.881, 0.0, 1.0, 0.949]" ], "question": "Where was the person currently performing the answer phone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1596, "options": [ "A. [0.64, 0.008, 1.0, 0.978]", "B. [0.872, 0.099, 1.0, 0.686]", "C. [0.001, 0.0, 0.422, 0.992]", "D. [0.429, 0.125, 0.936, 1.0]" ], "question": "Where was the person currently performing the text on/look at a cellphone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1687, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. sit, carry/hold (an object), listen to (a person)", "C. stand, listen to (a person), watch (a person)", "D. walk, carry/hold (an object)" ], "question": "What action is the person at the location [0.001, 0.051, 0.389, 0.984] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1698, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.0, 0.439, 0.991] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1719, "options": [ "A. [0.315, 0.0, 0.741, 0.874]", "B. [0.857, 0.0, 1.0, 0.913]", "C. [0.579, 0.0, 0.955, 0.996]", "D. [0.542, 0.422, 0.616, 0.762]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1751, "options": [ "A. [0.428, 0.019, 0.804, 1.0]", "B. [0.04, 0.0, 0.351, 1.0]", "C. [0.594, 0.018, 0.999, 1.0]", "D. [0.891, 0.0, 1.0, 0.903]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1751 ]

[ { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 925, "options": [ "A. watch (e.g., TV)", "B. walk", "C. carry/hold (an object) and talk to (e.g., self, a person, a group)", "D. hit (an object)" ], "question": "What action is the person currently in the [0.522, 0.077, 0.817, 0.97] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 984, "options": [ "A. bend/bow (at the waist) [0.399, 0.053, 0.755, 0.977] -> hug (a person) [0.399, 0.053, 0.755, 0.977] -> talk to (e.g., self, a person, a group) [0.399, 0.053, 0.755, 0.977]", "B. talk to (e.g., self, a person, a group) [0.282, 0.106, 0.845, 0.875] -> touch (an object) [0.215, 0.000, 0.858, 1.000] -> hug (a person) [0.539, 0.004, 0.813, 0.953]", "C. watch (e.g., TV) [0.550, 0.000, 0.880, 1.000] -> hug (a person) [0.227, 0.047, 0.826, 0.895] -> talk to (e.g., self, a person, a group) [0.288, 0.047, 0.885, 0.972]", "D. hug (a person) [0.341, 0.164, 0.770, 1.000] -> talk to (e.g., self, a person, a group) [0.599, 0.000, 0.570, 1.000] -> close (e.g., a door, a box) [0.383, 0.236, 0.885, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.399, 0.053, 0.755, 0.977] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 989, "options": [ "A. point to (an object) [0.263, 0.142, 0.791, 0.802] -> hug (a person) [0.481, 0.254, 0.733, 1.000] -> talk to (e.g., self, a person, a group) [0.373, 0.000, 0.731, 0.930]", "B. bend/bow (at the waist) [0.240, 0.027, 0.825, 0.999] -> hug (a person) [0.332, 0.059, 0.634, 0.857] -> extract [0.255, 0.038, 0.734, 0.847]", "C. talk to (e.g., self, a person, a group) [0.267, 0.112, 0.902, 1.000] -> hug (a person) [0.494, 0.057, 0.956, 0.969] -> martial art [0.540, 0.000, 0.915, 0.884]", "D. hug (a person) [0.379, 0.093, 0.762, 0.966] -> talk to (e.g., self, a person, a group) [0.379, 0.093, 0.762, 0.966] -> bend/bow (at the waist) [0.382, 0.095, 0.748, 0.963]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.382, 0.095, 0.748, 0.963] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 990, "options": [ "A. take a photo [0.343, 0.176, 0.745, 1.000] -> bend/bow (at the waist) [0.497, 0.227, 0.934, 1.000] -> talk to (e.g., self, a person, a group) [0.481, 0.069, 0.925, 0.935]", "B. brush teeth [0.323, 0.241, 0.661, 0.983] -> bend/bow (at the waist) [0.318, 0.184, 0.754, 0.946] -> hug (a person) [0.553, 0.000, 0.631, 1.000]", "C. hug (a person) [0.499, 0.040, 0.604, 0.809] -> hug (a person) [0.317, 0.057, 0.867, 1.000] -> bend/bow (at the waist) [0.190, 0.000, 0.730, 1.000]", "D. hug (a person) [0.379, 0.093, 0.762, 0.966] -> talk to (e.g., self, a person, a group) [0.379, 0.093, 0.762, 0.966] -> bend/bow (at the waist) [0.382, 0.095, 0.748, 0.963]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.382, 0.095, 0.748, 0.963] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1017, "options": [ "A. [0.364, 0.0, 0.667, 0.841]", "B. [0.019, 0.369, 0.124, 0.893]", "C. [0.671, 0.272, 0.94, 1.0]", "D. [0.462, 0.046, 0.96, 0.966]" ], "question": "Where is the person currently performing the answer phone located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1052, "options": [ "A. talk to (e.g., self, a person, a group) -> lie/sleep", "B. work on a computer -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> drink", "D. give/serve (an object) to (a person) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1063, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.201, 0.015, 0.942, 0.977] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1066, "options": [ "A. point to (an object) -> lie/sleep", "B. talk to (e.g., self, a person, a group) -> lie/sleep", "C. lie/sleep -> dig", "D. lie/sleep -> shovel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1084, "options": [ "A. [0.139, 0.068, 0.859, 0.973]", "B. [0.332, 0.343, 0.487, 0.983]", "C. [0.27, 0.316, 0.856, 1.0]", "D. [0.387, 0.339, 1.0, 0.695]" ], "question": "Where was the person currently performing the lie/sleep in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1116, "options": [ "A. [0.381, 0.498, 0.545, 0.824]", "B. [0.214, 0.0, 0.929, 0.981]", "C. [0.818, 0.365, 0.877, 0.969]", "D. [0.579, 0.29, 0.715, 0.982]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1160, "options": [ "A. [0.219, 0.049, 0.709, 0.783]", "B. [0.527, 0.274, 0.956, 1.0]", "C. [0.308, 0.181, 0.562, 0.993]", "D. [0.345, 0.011, 0.852, 0.984]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1190, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1224, "options": [ "A. [0.028, 0.008, 0.769, 0.99]", "B. [0.271, 0.0, 0.865, 1.0]", "C. [0.145, 0.0, 0.481, 0.748]", "D. [0.655, 0.311, 0.97, 0.978]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1251, "options": [ "A. walk, carry/hold (an object)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, touch (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.104, 0.329, 0.471, 0.966] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1260, "options": [ "A. talk to (e.g., self, a person, a group) -> fight/hit (a person)", "B. dig -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> dress/put on clothing" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1319, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1390, "options": [ "A. [0.217, 0.446, 0.213, 1.0]", "B. [0.332, 0.343, 0.487, 0.983]", "C. [0.0, 0.0, 0.742, 0.982]", "D. [0.557, 0.118, 0.615, 0.964]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1391, "options": [ "A. walk -> carry/hold (an object)", "B. stir -> carry/hold (an object)", "C. dig -> carry/hold (an object)", "D. climb (e.g., a mountain) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1444, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.197, 0.012, 0.95, 0.976] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1465, "options": [ "A. [0.597, 0.359, 0.713, 0.661]", "B. [0.332, 0.163, 0.622, 0.882]", "C. [0.782, 0.476, 0.919, 0.847]", "D. [0.818, 0.089, 0.825, 0.889]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1519, "options": [ "A. carry/hold (an object) -> walk", "B. text on/look at a cellphone -> carry/hold (an object)", "C. carry/hold (an object) -> text on/look at a cellphone", "D. watch (e.g., TV) -> text on/look at a cellphone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1528, "options": [ "A. carry/hold (an object) [0.254, 0.087, 0.790, 0.952] -> walk [0.371, 0.416, 0.385, 0.995] -> dress/put on clothing [0.461, 0.180, 0.604, 1.000]", "B. answer phone [0.385, 0.256, 0.598, 0.992] -> carry/hold (an object) [0.385, 0.256, 0.598, 0.992] -> walk [0.175, 0.341, 0.399, 1.0]", "C. walk [0.367, 0.249, 0.304, 1.000] -> push (an object) [0.358, 0.267, 0.586, 0.871] -> answer phone [0.213, 0.305, 0.447, 1.000]", "D. answer phone [0.301, 0.440, 0.734, 0.895] -> carry/hold (an object) [0.467, 0.255, 0.486, 1.000] -> write [0.176, 0.173, 0.321, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.175, 0.341, 0.399, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1531, "options": [ "A. carry/hold (an object) [0.019, 0.369, 0.124, 0.893] -> walk [0.082, 0.388, 0.148, 0.79] -> answer phone [0.082, 0.388, 0.148, 0.79]", "B. carry/hold (an object) [0.000, 0.175, 0.183, 0.886] -> extract [0.260, 0.565, 0.043, 0.974] -> answer phone [0.084, 0.449, 0.329, 0.643]", "C. answer phone [0.090, 0.438, 0.336, 0.812] -> walk [0.264, 0.305, 0.012, 0.638] -> grab (a person) [0.000, 0.417, 0.000, 0.953]", "D. dress/put on clothing [0.160, 0.477, 0.136, 0.866] -> answer phone [0.000, 0.432, 0.000, 0.740] -> carry/hold (an object) [0.000, 0.525, 0.123, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.082, 0.388, 0.148, 0.79] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1687, "options": [ "A. [0.361, 0.0, 1.0, 1.0]", "B. [0.308, 0.299, 0.985, 0.997]", "C. [0.187, 0.152, 0.806, 0.994]", "D. [0.043, 0.002, 0.659, 0.978]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1731, "options": [ "A. [0.0, 0.0, 0.931, 0.675]", "B. [0.0, 0.041, 0.667, 1.0]", "C. [0.622, 0.026, 0.989, 0.965]", "D. [0.18, 0.005, 0.883, 0.969]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1744, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1744 ]

[ { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 914, "options": [ "A. [0.243, 0.045, 0.666, 0.975]", "B. [0.131, 0.118, 0.18, 0.455]", "C. [0.289, 0.195, 0.365, 0.596]", "D. [0.086, 0.019, 0.608, 0.432]" ], "question": "Where was the person currently performing the answer phone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 925, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1021, "options": [ "A. [0.353, 0.391, 0.475, 0.977]", "B. [0.593, 0.433, 0.625, 1.0]", "C. [0.239, 0.036, 0.98, 0.975]", "D. [0.567, 0.106, 0.306, 0.777]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1137, "options": [ "A. [0.203, 0.02, 0.82, 1.0]", "B. [0.145, 0.168, 0.523, 1.0]", "C. [0.402, 0.334, 0.676, 0.992]", "D. [0.135, 0.516, 0.466, 1.0]" ], "question": "Where is the person currently performing the brush teeth located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1202, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.453, 0.312, 0.561, 0.841] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1336, "options": [ "A. stand, listen to (a person)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.103, 0.068, 0.79, 0.981] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1369, "options": [ "A. [0.527, 0.501, 0.895, 0.816]", "B. [0.796, 0.295, 0.959, 0.839]", "C. [0.701, 0.203, 0.849, 0.532]", "D. [0.55, 0.291, 0.662, 0.682]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1375, "options": [ "A. cut", "B. carry/hold (an object)", "C. answer phone", "D. dance" ], "question": "What action is the person currently in the [0.021, 0.279, 0.111, 0.591] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1420, "options": [ "A. [0.012, 0.277, 0.492, 0.901]", "B. [0.527, 0.027, 0.944, 0.997]", "C. [0.036, 0.018, 0.735, 0.985]", "D. [0.207, 0.107, 0.908, 0.893]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1464, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. stand, carry/hold (an object), listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.025, 0.31, 0.121, 0.617] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1530, "options": [ "A. [0.346, 0.24, 0.513, 0.989]", "B. [0.087, 0.265, 0.296, 0.978]", "C. [0.47, 0.125, 0.533, 0.817]", "D. [0.564, 0.336, 0.617, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1635, "options": [ "A. [0.052, 0.0, 0.199, 0.784]", "B. [0.0, 0.024, 0.747, 0.965]", "C. [0.069, 0.104, 0.339, 0.968]", "D. [0.36, 0.3, 0.533, 0.997]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1713, "options": [ "A. [0.408, 0.064, 0.897, 1.0]", "B. [0.163, 0.0, 0.791, 0.992]", "C. [0.726, 0.022, 1.0, 0.969]", "D. [0.0, 0.279, 0.851, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1744, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.819, 0.367, 1.0, 1.0] location in the current frame performing the eat?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1767, "options": [ "A. [0.582, 0.367, 0.951, 1.0]", "B. [0.515, 0.219, 0.749, 0.989]", "C. [0.658, 0.0, 0.886, 0.787]", "D. [0.756, 0.093, 1.0, 1.0]" ], "question": "What location in the frame is the person currently in the [0.485, 0.386, 0.578, 0.7] location likely to move to next?", "sub_answer_type": "Action Anticipation" } ]

[ 900, 1767 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 966, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.28, 0.074, 0.868, 1.0] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 994, "options": [ "A. 6", "B. 2", "C. 3", "D. 4" ], "question": "How many people in the current frame are performing the action: ride (e.g., a bike, a car, a horse)?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1027, "options": [ "A. ride (e.g., a bike, a car, a horse) [0.447, 0.131, 0.716, 1.0] -> talk to (e.g., self, a person, a group) [0.447, 0.131, 0.716, 1.0] -> carry/hold (an object) [0.479, 0.166, 0.713, 0.983]", "B. talk to (e.g., self, a person, a group) [0.571, 0.000, 0.626, 1.000] -> ride (e.g., a bike, a car, a horse) [0.426, 0.252, 0.855, 1.000] -> take (an object) from (a person) [0.645, 0.108, 0.690, 1.000]", "C. brush teeth [0.256, 0.000, 0.891, 1.000] -> talk to (e.g., self, a person, a group) [0.611, 0.000, 0.536, 0.971] -> carry/hold (an object) [0.358, 0.124, 0.704, 0.851]", "D. ride (e.g., a bike, a car, a horse) [0.357, 0.201, 0.561, 0.921] -> kick (a person) [0.636, 0.029, 0.548, 0.880] -> talk to (e.g., self, a person, a group) [0.328, 0.319, 0.795, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.479, 0.166, 0.713, 0.983] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1073, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1098, "options": [ "A. walk, carry/hold (an object)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, ride (e.g., a bike, a car, a horse), listen to (a person)", "D. walk, watch (a person)" ], "question": "What action is the person at the location [0.155, 0.22, 0.445, 0.983] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1219, "options": [ "A. [0.203, 0.298, 0.257, 1.0]", "B. [0.644, 0.348, 0.814, 0.921]", "C. [0.132, 0.008, 0.54, 0.999]", "D. [0.157, 0.0, 0.722, 0.707]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1237, "options": [ "A. walk", "B. open (e.g., a window, a car door)", "C. paint", "D. crouch/kneel" ], "question": "What action is the person currently in the [0.645, 0.026, 0.817, 0.975] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1280, "options": [ "A. [0.376, 0.297, 0.803, 1.0]", "B. [0.574, 0.044, 0.727, 0.941]", "C. [0.039, 0.0, 0.285, 0.927]", "D. [0.253, 0.151, 0.515, 0.932]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1294, "options": [ "A. [0.459, 0.483, 0.67, 0.999]", "B. [0.385, 0.209, 0.566, 0.985]", "C. [0.574, 0.042, 0.852, 0.998]", "D. [0.099, 0.0, 0.397, 0.788]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1315, "options": [ "A. carry/hold (an object) [0.144, 0.046, 0.562, 0.96] -> talk to (e.g., self, a person, a group) [0.144, 0.046, 0.562, 0.96] -> lie/sleep [0.144, 0.046, 0.562, 0.96]", "B. talk to (e.g., self, a person, a group) [0.129, 0.195, 0.480, 0.941] -> carry/hold (an object) [0.000, 0.243, 0.425, 0.955] -> get up [0.034, 0.160, 0.637, 0.776]", "C. talk to (e.g., self, a person, a group) [0.080, 0.000, 0.586, 0.925] -> grab (a person) [0.274, 0.108, 0.739, 0.899] -> carry/hold (an object) [0.037, 0.000, 0.670, 0.919]", "D. lie/sleep [0.267, 0.000, 0.386, 1.000] -> text on/look at a cellphone [0.195, 0.188, 0.654, 1.000] -> carry/hold (an object) [0.017, 0.002, 0.736, 0.980]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.144, 0.046, 0.562, 0.96] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1320, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1343, "options": [ "A. carry/hold (an object) [0.135, 0.118, 0.790, 0.984] -> bend/bow (at the waist) [0.297, 0.172, 0.581, 0.992] -> swim [0.072, 0.004, 0.571, 0.884]", "B. carry/hold (an object) [0.000, 0.000, 0.499, 0.878] -> carry/hold (an object) [0.000, 0.000, 0.448, 0.931] -> bend/bow (at the waist) [0.299, 0.000, 0.738, 0.951]", "C. bend/bow (at the waist) [0.132, 0.017, 0.62, 1.0] -> carry/hold (an object) [0.132, 0.017, 0.62, 1.0] -> talk to (e.g., self, a person, a group) [0.132, 0.017, 0.62, 1.0]", "D. text on/look at a cellphone [0.031, 0.000, 0.788, 1.000] -> bend/bow (at the waist) [0.175, 0.086, 0.495, 1.000] -> carry/hold (an object) [0.049, 0.070, 0.514, 0.935]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.132, 0.017, 0.62, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1350, "options": [ "A. stand, listen to (a person), watch (a person)", "B. sit, ride (e.g., a bike, a car, a horse)", "C. carry/hold (an object), lie/sleep", "D. stand, carry/hold (an object), smoke" ], "question": "What action is the person at the location [0.165, 0.132, 0.621, 0.945] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1384, "options": [ "A. lie/sleep [0.327, 0.086, 1.000, 0.856] -> carry/hold (an object) [0.513, 0.012, 0.770, 0.785] -> shoot [0.322, 0.009, 0.820, 1.000]", "B. talk to (e.g., self, a person, a group) [0.471, 0.000, 0.971, 0.816] -> carry/hold (an object) [0.449, 0.015, 0.662, 1.000] -> exit [0.559, 0.108, 0.933, 1.000]", "C. lie/sleep [0.261, 0.088, 0.811, 0.931] -> point to (an object) [0.300, 0.000, 0.778, 0.839] -> talk to (e.g., self, a person, a group) [0.426, 0.107, 0.944, 1.000]", "D. carry/hold (an object) [0.36, 0.0, 0.858, 0.965] -> talk to (e.g., self, a person, a group) [0.36, 0.0, 0.858, 0.965] -> lie/sleep [0.36, 0.0, 0.858, 0.965]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.36, 0.0, 0.858, 0.965] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1392, "options": [ "A. [0.379, 0.0, 0.89, 0.733]", "B. [0.554, 0.156, 0.864, 0.987]", "C. [0.226, 0.007, 0.833, 0.986]", "D. [0.133, 0.0, 0.572, 0.713]" ], "question": "Where was the person currently performing the lie/sleep in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1545, "options": [ "A. [0.121, 0.321, 0.383, 0.833]", "B. [0.193, 0.045, 0.591, 0.983]", "C. [0.526, 0.135, 0.7, 0.992]", "D. [0.0, 0.0, 0.341, 0.81]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1561, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1580, "options": [ "A. [0.571, 0.001, 0.656, 0.667]", "B. [0.144, 0.014, 0.702, 1.0]", "C. [0.0, 0.189, 0.434, 1.0]", "D. [0.008, 0.0, 0.436, 0.747]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1703, "options": [ "A. fall down [0.476, 0.000, 0.903, 1.000] -> talk to (e.g., self, a person, a group) [0.425, 0.094, 0.896, 0.775] -> carry/hold (an object) [0.499, 0.000, 0.886, 0.979]", "B. carry/hold (an object) [0.522, 0.126, 0.621, 0.775] -> martial art [0.411, 0.000, 0.852, 0.785] -> talk to (e.g., self, a person, a group) [0.691, 0.000, 0.624, 0.906]", "C. carry/hold (an object) [0.574, 0.044, 0.727, 0.941] -> talk to (e.g., self, a person, a group) [0.574, 0.044, 0.727, 0.941] -> walk [0.536, 0.0, 0.853, 0.961]", "D. walk [0.567, 0.000, 0.747, 0.957] -> put down [0.610, 0.000, 0.849, 1.000] -> carry/hold (an object) [0.474, 0.000, 0.729, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.536, 0.0, 0.853, 0.961] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1713, "options": [ "A. [0.0, 0.244, 0.447, 0.958]", "B. [0.664, 0.311, 0.853, 0.919]", "C. [0.237, 0.227, 0.585, 0.952]", "D. [0.492, 0.0, 0.669, 1.0]" ], "question": "Where was the person currently performing the bend/bow (at the waist) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1713 ]

[ { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1036, "options": [ "A. 5", "B. 2", "C. 3", "D. 4" ], "question": "How many people in the current frame are performing the action: martial art?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1065, "options": [ "A. stand, watch (a person)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. fight/hit (a person), martial art", "D. stand, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.722, 0.488, 0.835, 0.886] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1068, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1096, "options": [ "A. [0.319, 0.256, 0.605, 0.903]", "B. [0.0, 0.0, 0.398, 0.776]", "C. [0.004, 0.14, 0.57, 0.906]", "D. [0.0, 0.296, 0.271, 0.966]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1117, "options": [ "A. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "B. extract -> carry/hold (an object)", "C. press -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> fall down" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1256, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1259, "options": [ "A. [0.366, 0.445, 0.392, 0.937]", "B. [0.153, 0.228, 0.425, 0.883]", "C. [0.089, 0.0, 0.151, 0.694]", "D. [0.711, 0.281, 0.974, 0.9]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1286, "options": [ "A. carry/hold (an object) -> throw", "B. talk to (e.g., self, a person, a group) -> shoot", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> work on a computer" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1306, "options": [ "A. [0.235, 0.445, 1.0, 1.0]", "B. [0.717, 0.2, 0.577, 0.599]", "C. [0.105, 0.335, 0.773, 0.666]", "D. [0.488, 0.24, 0.875, 0.888]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1322, "options": [ "A. carry/hold (an object)", "B. point to (an object)", "C. talk to (e.g., self, a person, a group)", "D. fight/hit (a person)" ], "question": "What action is the person currently in the [0.678, 0.276, 0.901, 0.894] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1412, "options": [ "A. [0.078, 0.226, 0.408, 1.0]", "B. [0.11, 0.104, 0.692, 0.904]", "C. [0.388, 0.396, 0.579, 0.881]", "D. [0.216, 0.153, 0.897, 0.846]" ], "question": "Where is the person currently performing the martial art located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1490, "options": [ "A. talk to (e.g., self, a person, a group) -> write", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. play board game -> talk to (e.g., self, a person, a group)", "D. open (e.g., a window, a car door) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1528, "options": [ "A. 2", "B. 3", "C. 9", "D. 4" ], "question": "How many people in the current frame are performing the action: fight/hit (a person)?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1663, "options": [ "A. 6", "B. 3", "C. 2", "D. 4" ], "question": "How many people in the current frame are performing the action: martial art?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1667, "options": [ "A. 9", "B. 5", "C. 7", "D. 8" ], "question": "How many people were performing the watch (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1679, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.07, 0.077, 0.756, 0.878] location in the current frame performing the lie/sleep?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1680, "options": [ "A. hand clap -> lie/sleep", "B. lie/sleep -> work on a computer", "C. fight/hit (a person) -> lie/sleep", "D. cut -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1699, "options": [ "A. fall down -> get up", "B. lie/sleep -> smoke", "C. fall down -> brush teeth", "D. lie/sleep -> fall down" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1746, "options": [ "A. [0.359, 0.054, 0.657, 0.626]", "B. [0.189, 0.027, 0.45, 0.867]", "C. [0.433, 0.23, 0.877, 0.894]", "D. [0.297, 0.224, 0.606, 0.897]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 916, 1746 ]

[ { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 913, "options": [ "A. walk and carry/hold (an object)", "B. play board game", "C. talk to (e.g., self, a person, a group) and lie/sleep", "D. exit" ], "question": "What action is the person currently in the [0.433, 0.119, 0.709, 0.984] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 925, "options": [ "A. extract [0.407, 0.000, 0.410, 0.767] -> carry/hold (an object) [0.527, 0.004, 0.758, 0.797] -> talk to (e.g., self, a person, a group) [0.343, 0.226, 0.546, 0.783]", "B. carry/hold (an object) [0.424, 0.063, 0.609, 0.739] -> close (e.g., a door, a box) [0.424, 0.063, 0.609, 0.739] -> talk to (e.g., self, a person, a group) [0.424, 0.063, 0.609, 0.739]", "C. carry/hold (an object) [0.431, 0.251, 0.667, 0.734] -> talk to (e.g., self, a person, a group) [0.268, 0.012, 0.417, 0.691] -> point to (an object) [0.551, 0.000, 0.684, 0.896]", "D. close (e.g., a door, a box) [0.313, 0.024, 0.659, 0.858] -> ride (e.g., a bike, a car, a horse) [0.556, 0.253, 0.621, 0.648] -> talk to (e.g., self, a person, a group) [0.257, 0.040, 0.731, 0.769]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.424, 0.063, 0.609, 0.739] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 934, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 963, "options": [ "A. sit, carry/hold (an object), write", "B. sit, listen to (a person)", "C. sit, listen to (a person), watch (a person)", "D. bend/bow (at the waist), pull (an object)" ], "question": "What action is the person at the location [0.062, 0.072, 0.381, 0.743] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 975, "options": [ "A. talk to (e.g., self, a person, a group) -> lie/sleep", "B. talk to (e.g., self, a person, a group) -> jump/leap", "C. lie/sleep -> paint", "D. lie/sleep -> clink glass" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1024, "options": [ "A. put down -> lie/sleep -> carry/hold (an object)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> lie/sleep", "C. talk to (e.g., self, a person, a group) -> carry/hold (an object) -> sing to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> work on a computer -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1094, "options": [ "A. carry/hold (an object) -> push (another person)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. throw -> carry/hold (an object)", "D. talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1136, "options": [ "A. [0.123, 0.0, 1.0, 0.703]", "B. [0.206, 0.332, 0.53, 1.0]", "C. [0.505, 0.253, 0.957, 0.72]", "D. [0.223, 0.089, 0.947, 0.988]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1144, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.332, 0.146, 0.797, 0.994] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1185, "options": [ "A. work on a computer -> open (e.g., a window, a car door)", "B. walk -> open (e.g., a window, a car door)", "C. drive (e.g., a car, a truck) -> walk", "D. walk -> grab (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1222, "options": [ "A. [0.279, 0.0, 0.751, 0.752]", "B. [0.433, 0.266, 0.566, 1.0]", "C. [0.65, 0.154, 0.971, 0.997]", "D. [0.461, 0.0, 0.805, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1296, "options": [ "A. walk [0.539, 0.11, 0.783, 0.916] -> carry/hold (an object) [0.539, 0.11, 0.783, 0.916] -> talk to (e.g., self, a person, a group) [0.539, 0.11, 0.783, 0.916]", "B. carry/hold (an object) [0.556, 0.094, 0.711, 0.955] -> dance [0.364, 0.000, 0.586, 0.782] -> talk to (e.g., self, a person, a group) [0.606, 0.000, 0.605, 0.853]", "C. crouch/kneel [0.705, 0.233, 0.745, 0.751] -> talk to (e.g., self, a person, a group) [0.663, 0.000, 0.635, 1.000] -> carry/hold (an object) [0.487, 0.111, 0.751, 0.722]", "D. walk [0.376, 0.005, 0.823, 0.930] -> carry/hold (an object) [0.386, 0.061, 0.893, 0.978] -> push (another person) [0.427, 0.300, 0.882, 0.951]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.539, 0.11, 0.783, 0.916] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1373, "options": [ "A. talk to (e.g., self, a person, a group) -> take (an object) from (a person)", "B. take (an object) from (a person) -> push (an object)", "C. play board game -> take (an object) from (a person)", "D. talk to (e.g., self, a person, a group) -> give/serve (an object) to (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1383, "options": [ "A. 4", "B. 6", "C. 5", "D. 3" ], "question": "How many people were performing the carry/hold (an object) in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1438, "options": [ "A. talk to (e.g., self, a person, a group) [0.178, 0.000, 0.494, 1.000] -> walk [0.000, 0.066, 0.562, 1.000] -> kiss (a person) [0.219, 0.000, 0.541, 0.954]", "B. talk to (e.g., self, a person, a group) [0.274, 0.083, 0.468, 0.889] -> walk [0.000, 0.176, 0.414, 1.000] -> play with pets [0.198, 0.000, 0.522, 1.000]", "C. walk [0.087, 0.011, 0.366, 0.961] -> carry/hold (an object) [0.087, 0.011, 0.366, 0.961] -> talk to (e.g., self, a person, a group) [0.087, 0.011, 0.366, 0.961]", "D. walk [0.108, 0.103, 0.559, 0.977] -> fall down [0.000, 0.060, 0.450, 0.763] -> talk to (e.g., self, a person, a group) [0.000, 0.130, 0.179, 0.960]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.087, 0.011, 0.366, 0.961] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1448, "options": [ "A. [0.246, 0.337, 0.424, 1.0]", "B. [0.047, 0.067, 0.495, 0.992]", "C. [0.366, 0.056, 0.741, 0.977]", "D. [0.044, 0.281, 0.199, 1.0]" ], "question": "Where is the person currently performing the dress/put on clothing located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1465, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.269, 0.009, 0.768, 0.971] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1466, "options": [ "A. smoke [0.439, 0.022, 0.877, 1.000] -> talk to (e.g., self, a person, a group) [0.391, 0.000, 0.711, 1.000] -> cook [0.406, 0.000, 0.889, 0.883] -> carry/hold (an object) [0.232, 0.000, 0.740, 0.955]", "B. carry/hold (an object) [0.119, 0.000, 0.803, 0.774] -> eat [0.534, 0.237, 0.948, 0.856] -> smoke [0.334, 0.204, 0.758, 0.840] -> walk [0.281, 0.000, 0.847, 0.906]", "C. smoke [0.349, 0.043, 0.898, 0.976] -> talk to (e.g., self, a person, a group) [0.349, 0.043, 0.898, 0.976] -> walk [0.269, 0.009, 0.768, 0.971] -> carry/hold (an object) [0.269, 0.009, 0.768, 0.971]", "D. walk [0.439, 0.190, 0.847, 1.000] -> smoke [0.271, 0.115, 0.740, 1.000] -> talk to (e.g., self, a person, a group) [0.327, 0.086, 0.984, 1.000] -> lift (a person) [0.326, 0.064, 0.575, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.269, 0.009, 0.768, 0.971] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1483, "options": [ "A. 7", "B. 4", "C. 3", "D. 5" ], "question": "How many people were performing the watch (a person) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1524, "options": [ "A. stand, listen to (a person)", "B. walk, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, drink", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.039, 0.319, 0.304, 0.967] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1615, "options": [ "A. 7", "B. 4", "C. 2", "D. 3" ], "question": "How many people in the current frame are performing the action: talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1634, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.096, 0.05, 0.905, 0.976] location in the current frame performing the ride (e.g., a bike, a car, a horse)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1675, "options": [ "A. [0.198, 0.619, 0.446, 0.997]", "B. [0.11, 0.782, 0.192, 1.0]", "C. [0.4, 0.329, 0.165, 0.708]", "D. [0.037, 0.352, 0.393, 0.988]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1755, "options": [ "A. fight/hit (a person) [0.420, 0.152, 0.768, 1.000] -> talk to (e.g., self, a person, a group) [0.455, 0.377, 0.538, 0.869] -> carry/hold (an object) [0.573, 0.180, 0.760, 0.976]", "B. carry/hold (an object) [0.436, 0.425, 0.618, 1.000] -> give/serve (an object) to (a person) [0.295, 0.048, 0.769, 1.000] -> walk [0.196, 0.401, 0.479, 0.870]", "C. carry/hold (an object) [0.606, 0.187, 0.695, 1.000] -> point to (an object) [0.466, 0.346, 0.545, 0.987] -> talk to (e.g., self, a person, a group) [0.269, 0.398, 0.859, 1.000]", "D. carry/hold (an object) [0.413, 0.238, 0.691, 0.992] -> talk to (e.g., self, a person, a group) [0.413, 0.238, 0.691, 0.992] -> walk [0.376, 0.222, 0.612, 0.969]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.376, 0.222, 0.612, 0.969] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1758, "options": [ "A. [0.0, 0.0, 0.49, 0.744]", "B. [0.094, 0.284, 0.182, 1.0]", "C. [0.263, 0.037, 0.793, 0.953]", "D. [0.11, 0.052, 0.478, 0.994]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1758 ]

[ { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 928, "options": [ "A. get up -> carry/hold (an object) -> walk", "B. walk -> run/jog -> drink", "C. walk -> carry/hold (an object) -> drink", "D. walk -> play with pets -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 983, "options": [ "A. carry/hold (an object) [0.304, 0.153, 0.69, 0.98] -> crouch/kneel [0.304, 0.153, 0.69, 0.98] -> talk to (e.g., self, a person, a group) [0.304, 0.153, 0.69, 0.98]", "B. push (an object) [0.276, 0.066, 0.811, 1.000] -> crouch/kneel [0.388, 0.219, 0.528, 1.000] -> talk to (e.g., self, a person, a group) [0.353, 0.067, 0.877, 1.000]", "C. carry/hold (an object) [0.419, 0.298, 0.849, 0.919] -> read [0.356, 0.160, 0.739, 0.910] -> crouch/kneel [0.464, 0.242, 0.736, 0.916]", "D. carry/hold (an object) [0.293, 0.185, 0.816, 0.849] -> talk to (e.g., self, a person, a group) [0.461, 0.192, 0.566, 0.820] -> write [0.430, 0.251, 0.738, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.304, 0.153, 0.69, 0.98] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 988, "options": [ "A. fishing [0.387, 0.295, 0.701, 0.917] -> talk to (e.g., self, a person, a group) [0.268, 0.240, 0.621, 1.000] -> crouch/kneel [0.434, 0.275, 0.518, 1.000]", "B. carry/hold (an object) [0.304, 0.153, 0.69, 0.98] -> crouch/kneel [0.304, 0.153, 0.69, 0.98] -> talk to (e.g., self, a person, a group) [0.304, 0.153, 0.69, 0.98]", "C. carry/hold (an object) [0.144, 0.225, 0.639, 0.953] -> crouch/kneel [0.449, 0.275, 0.675, 0.899] -> hug (a person) [0.376, 0.126, 0.875, 0.904]", "D. crouch/kneel [0.115, 0.012, 0.868, 1.000] -> dress/put on clothing [0.321, 0.307, 0.707, 1.000] -> carry/hold (an object) [0.223, 0.113, 0.616, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.304, 0.153, 0.69, 0.98] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 991, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 993, "options": [ "A. [0.321, 0.337, 0.408, 1.0]", "B. [0.001, 0.0, 0.488, 0.81]", "C. [0.28, 0.156, 0.699, 0.982]", "D. [0.523, 0.237, 0.848, 0.905]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1016, "options": [ "A. carry/hold (an object) -> take (an object) from (a person) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> get up -> carry/hold (an object)", "C. carry/hold (an object) -> push (another person) -> take (an object) from (a person)", "D. read -> take (an object) from (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1037, "options": [ "A. [0.317, 0.0, 0.555, 0.686]", "B. [0.43, 0.177, 0.82, 0.969]", "C. [0.12, 0.012, 0.668, 0.982]", "D. [0.0, 0.288, 0.552, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1063, "options": [ "A. [0.392, 0.221, 0.61, 0.848]", "B. [0.096, 0.418, 0.835, 1.0]", "C. [0.674, 0.05, 0.539, 0.622]", "D. [0.594, 0.391, 0.748, 0.924]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1172, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.124, 0.084, 0.512, 0.981] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1194, "options": [ "A. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "B. talk to (e.g., self, a person, a group) -> point to (an object)", "C. shovel -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> hand shake" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1258, "options": [ "A. [0.623, 0.167, 0.831, 0.97]", "B. [0.535, 0.028, 0.629, 1.0]", "C. [0.423, 0.014, 0.428, 0.666]", "D. [0.333, 0.193, 0.471, 0.893]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1283, "options": [ "A. catch (an object) -> read", "B. point to (an object) -> carry/hold (an object)", "C. read -> carry/hold (an object)", "D. carry/hold (an object) -> hand wave" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1354, "options": [ "A. [0.431, 0.141, 0.755, 0.975]", "B. [0.492, 0.011, 0.516, 0.694]", "C. [0.347, 0.223, 0.433, 0.784]", "D. [0.377, 0.4, 0.96, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1357, "options": [ "A. play with pets -> carry/hold (an object)", "B. carry/hold (an object) -> play board game", "C. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "D. hug (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1407, "options": [ "A. [0.48, 0.0, 0.715, 0.77]", "B. [0.243, 0.115, 0.479, 0.9]", "C. [0.554, 0.071, 0.879, 0.935]", "D. [0.235, 0.325, 0.354, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1447, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1533, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1622, "options": [ "A. sit, listen to (a person)", "B. stand, talk to (e.g., self, a person, a group)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.133, 0.039, 0.682, 0.976] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1657, "options": [ "A. stand, carry/hold (an object)", "B. stand, carry/hold (an object), watch (a person)", "C. walk, watch (a person)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.385, 0.145, 0.491, 0.808] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1660, "options": [ "A. carry/hold (an object)", "B. touch (an object)", "C. cook", "D. chop" ], "question": "What action is the person currently in the [0.578, 0.387, 0.746, 0.916] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1714, "options": [ "A. walk [0.535, 0.091, 0.461, 1.000] -> talk to (e.g., self, a person, a group) [0.257, 0.331, 0.594, 0.789] -> hand wave [0.307, 0.289, 0.670, 1.000]", "B. play with kids [0.297, 0.385, 0.685, 0.905] -> carry/hold (an object) [0.368, 0.114, 0.459, 0.983] -> talk to (e.g., self, a person, a group) [0.483, 0.296, 0.713, 1.000]", "C. walk [0.465, 0.134, 0.506, 1.000] -> talk to (e.g., self, a person, a group) [0.322, 0.244, 0.578, 0.842] -> push (an object) [0.272, 0.114, 0.353, 0.855]", "D. carry/hold (an object) [0.351, 0.211, 0.548, 0.966] -> talk to (e.g., self, a person, a group) [0.351, 0.211, 0.548, 0.966] -> walk [0.367, 0.191, 0.557, 0.965]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.367, 0.191, 0.557, 0.965] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1722, "options": [ "A. carry/hold (an object) [0.352, 0.225, 0.58, 0.838] -> put down [0.352, 0.225, 0.58, 0.838] -> talk to (e.g., self, a person, a group) [0.352, 0.225, 0.58, 0.838]", "B. carry/hold (an object) [0.338, 0.297, 0.510, 0.893] -> talk to (e.g., self, a person, a group) [0.232, 0.398, 0.479, 0.666] -> dress/put on clothing [0.203, 0.363, 0.709, 0.852]", "C. talk to (e.g., self, a person, a group) [0.311, 0.041, 0.616, 0.794] -> play board game [0.219, 0.309, 0.534, 0.833] -> put down [0.401, 0.236, 0.406, 0.997]", "D. put down [0.156, 0.333, 0.500, 0.934] -> put down [0.489, 0.279, 0.708, 0.885] -> talk to (e.g., self, a person, a group) [0.305, 0.227, 0.655, 0.973]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.352, 0.225, 0.58, 0.838] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1773, "options": [ "A. talk to (e.g., self, a person, a group) [0.126, 0.495, 0.651, 0.987] -> exit [0.231, 0.526, 0.427, 1.000] -> lie/sleep [0.148, 0.481, 0.705, 0.992]", "B. carry/hold (an object) [0.272, 0.312, 0.515, 0.879] -> talk to (e.g., self, a person, a group) [0.024, 0.539, 0.380, 0.804] -> ride (e.g., a bike, a car, a horse) [0.000, 0.630, 0.407, 0.796]", "C. shovel [0.315, 0.352, 0.556, 0.781] -> talk to (e.g., self, a person, a group) [0.280, 0.438, 0.454, 0.859] -> carry/hold (an object) [0.156, 0.406, 0.604, 0.992]", "D. lie/sleep [0.188, 0.437, 0.545, 0.976] -> carry/hold (an object) [0.199, 0.408, 0.55, 0.984] -> talk to (e.g., self, a person, a group) [0.199, 0.408, 0.55, 0.984]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.199, 0.408, 0.55, 0.984] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1792, "options": [ "A. [0.122, 0.49, 0.164, 0.932]", "B. [0.0, 0.099, 0.539, 0.496]", "C. [0.169, 0.216, 0.419, 0.735]", "D. [0.139, 0.127, 0.712, 0.984]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1792 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 934, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, listen to (a person)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.012, 0.108, 0.468, 0.952] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 963, "options": [ "A. carry/hold (an object) [0.460, 0.296, 1.000, 1.000] -> get up [0.630, 0.005, 1.000, 0.802] -> write [0.430, 0.030, 0.797, 0.981]", "B. talk to (e.g., self, a person, a group) [0.282, 0.224, 1.000, 0.839] -> row boat [0.601, 0.239, 0.914, 0.969] -> carry/hold (an object) [0.320, 0.319, 1.000, 0.829]", "C. write [0.479, 0.139, 0.974, 0.92] -> talk to (e.g., self, a person, a group) [0.479, 0.139, 0.974, 0.92] -> carry/hold (an object) [0.463, 0.133, 0.982, 0.917]", "D. throw [0.349, 0.245, 0.929, 0.970] -> carry/hold (an object) [0.563, 0.000, 0.932, 0.969] -> talk to (e.g., self, a person, a group) [0.524, 0.226, 1.000, 0.752]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.463, 0.133, 0.982, 0.917] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 987, "options": [ "A. carry/hold (an object) -> write -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> cut -> write", "C. write -> climb (e.g., a mountain) -> carry/hold (an object)", "D. drink -> write -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1030, "options": [ "A. [0.176, 0.291, 0.482, 0.891]", "B. [0.253, 0.494, 0.397, 1.0]", "C. [0.343, 0.215, 0.698, 0.867]", "D. [0.558, 0.148, 0.984, 0.912]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1060, "options": [ "A. carry/hold (an object) -> sail boat", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. give/serve (an object) to (a person) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> open (e.g., a window, a car door)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1072, "options": [ "A. sail boat [0.274, 0.363, 0.873, 0.752] -> talk to (e.g., self, a person, a group) [0.430, 0.158, 1.000, 0.707] -> pull (an object) [0.271, 0.091, 0.792, 0.690]", "B. hand wave [0.543, 0.092, 0.719, 0.786] -> pull (an object) [0.344, 0.335, 0.728, 0.968] -> talk to (e.g., self, a person, a group) [0.223, 0.275, 0.902, 0.988]", "C. carry/hold (an object) [0.363, 0.201, 0.917, 0.868] -> pull (an object) [0.363, 0.201, 0.917, 0.868] -> talk to (e.g., self, a person, a group) [0.363, 0.201, 0.917, 0.868]", "D. pull (an object) [0.200, 0.297, 0.911, 0.709] -> carry/hold (an object) [0.546, 0.008, 1.000, 0.719] -> talk to (e.g., self, a person, a group) [0.251, 0.229, 1.000, 0.803]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.363, 0.201, 0.917, 0.868] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1075, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1098, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, carry/hold (an object), watch (e.g., TV)", "C. watch (a person), martial art", "D. sit, carry/hold (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.32, 0.226, 0.538, 0.564] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1175, "options": [ "A. clink glass -> lie/sleep", "B. lie/sleep -> get up", "C. lie/sleep -> hand wave", "D. lie/sleep -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1192, "options": [ "A. [0.049, 0.095, 0.617, 0.613]", "B. [0.439, 0.0, 0.727, 0.594]", "C. [0.481, 0.173, 0.926, 0.887]", "D. [0.228, 0.241, 0.459, 0.783]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1226, "options": [ "A. talk to (e.g., self, a person, a group) [0.236, 0.089, 0.692, 0.896] -> carry/hold (an object) [0.492, 0.322, 0.449, 1.000] -> dress/put on clothing [0.530, 0.212, 0.341, 0.768]", "B. talk to (e.g., self, a person, a group) [0.497, 0.280, 0.340, 0.844] -> talk to (e.g., self, a person, a group) [0.176, 0.137, 0.327, 0.917] -> walk [0.507, 0.262, 0.692, 0.874]", "C. walk [0.359, 0.263, 0.517, 0.867] -> carry/hold (an object) [0.359, 0.263, 0.517, 0.867] -> talk to (e.g., self, a person, a group) [0.359, 0.263, 0.517, 0.867]", "D. carry/hold (an object) [0.214, 0.452, 0.712, 0.880] -> talk to (e.g., self, a person, a group) [0.384, 0.151, 0.472, 1.000] -> kick (an object) [0.182, 0.185, 0.458, 0.898]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.359, 0.263, 0.517, 0.867] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1234, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.037, 0.138, 0.602, 0.86] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1280, "options": [ "A. stand, listen to (a person), watch (a person)", "B. sit, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.0, 0.113, 0.271, 0.913] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1290, "options": [ "A. carry/hold (an object) -> hand clap", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. hug (a person) -> talk to (e.g., self, a person, a group)", "D. lift/pick up -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1313, "options": [ "A. talk to (e.g., self, a person, a group) [0.356, 0.384, 0.880, 1.000] -> bend/bow (at the waist) [0.705, 0.225, 0.687, 0.775] -> shovel [0.482, 0.364, 0.690, 0.873]", "B. bend/bow (at the waist) [0.434, 0.101, 0.696, 0.684] -> talk to (e.g., self, a person, a group) [0.536, 0.412, 0.981, 0.676] -> open (e.g., a window, a car door) [0.497, 0.192, 0.699, 0.662]", "C. bend/bow (at the waist) [0.399, 0.355, 0.732, 0.805] -> touch (an object) [0.556, 0.386, 0.762, 0.794] -> talk to (e.g., self, a person, a group) [0.697, 0.135, 0.648, 1.000]", "D. bend/bow (at the waist) [0.526, 0.263, 0.823, 0.851] -> touch (an object) [0.526, 0.263, 0.823, 0.851] -> talk to (e.g., self, a person, a group) [0.526, 0.263, 0.823, 0.851]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.526, 0.263, 0.823, 0.851] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1379, "options": [ "A. walk [0.750, 0.024, 0.952, 0.732] -> swim [0.682, 0.376, 0.848, 0.716] -> talk to (e.g., self, a person, a group) [0.530, 0.173, 1.000, 0.855]", "B. walk [0.568, 0.19, 0.947, 0.901] -> carry/hold (an object) [0.568, 0.19, 0.947, 0.901] -> talk to (e.g., self, a person, a group) [0.568, 0.19, 0.947, 0.901]", "C. walk [0.533, 0.382, 0.926, 1.000] -> sail boat [0.744, 0.298, 0.761, 1.000] -> carry/hold (an object) [0.378, 0.296, 1.000, 0.807]", "D. talk to (e.g., self, a person, a group) [0.599, 0.303, 0.870, 0.898] -> dance [0.496, 0.114, 0.980, 0.722] -> carry/hold (an object) [0.688, 0.253, 0.954, 0.880]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.568, 0.19, 0.947, 0.901] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1456, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1505, "options": [ "A. sing to (e.g., self, a person, a group)", "B. carry/hold (an object) and climb (e.g., a mountain)", "C. kick (a person)", "D. push (another person)" ], "question": "What action is the person currently in the [0.081, 0.228, 0.52, 0.899] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1509, "options": [ "A. [0.389, 0.327, 0.837, 0.898]", "B. [0.289, 0.549, 0.978, 1.0]", "C. [0.0, 0.327, 0.549, 0.877]", "D. [0.12, 0.304, 0.769, 0.881]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1554, "options": [ "A. [0.456, 0.161, 0.984, 0.89]", "B. [0.227, 0.0, 1.0, 0.651]", "C. [0.754, 0.357, 0.906, 1.0]", "D. [0.283, 0.234, 0.545, 0.837]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1608, "options": [ "A. bend/bow (at the waist) -> row boat", "B. walk -> bend/bow (at the waist)", "C. bend/bow (at the waist) -> carry/hold (an object)", "D. bend/bow (at the waist) -> climb (e.g., a mountain)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1734, "options": [ "A. sit, carry/hold (an object), listen to (a person), watch (a person)", "B. stand, carry/hold (an object), listen to (a person)", "C. stand, talk to (e.g., self, a person, a group)", "D. stand, carry/hold (an object)" ], "question": "What action is the person at the location [0.119, 0.194, 0.368, 0.88] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1748, "options": [ "A. [0.909, 0.193, 0.886, 0.814]", "B. [0.406, 0.103, 1.0, 0.946]", "C. [0.645, 0.392, 0.968, 0.892]", "D. [0.08, 0.251, 0.345, 0.849]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1764, "options": [ "A. [0.251, 0.09, 0.923, 0.989]", "B. [0.0, 0.0, 0.022, 0.834]", "C. [0.001, 0.13, 0.289, 0.892]", "D. [0.234, 0.0, 0.511, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1764 ]

[ { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 912, "options": [ "A. [0.51, 0.244, 0.87, 0.971]", "B. [0.725, 0.462, 0.973, 1.0]", "C. [0.791, 0.031, 1.0, 1.0]", "D. [0.067, 0.115, 0.921, 0.978]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 918, "options": [ "A. smoke -> talk to (e.g., self, a person, a group)", "B. smoke -> work on a computer", "C. shoot -> smoke", "D. talk to (e.g., self, a person, a group) -> clink glass" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 960, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 979, "options": [ "A. [0.024, 0.065, 0.861, 0.98]", "B. [0.375, 0.025, 1.0, 0.769]", "C. [0.489, 0.149, 1.0, 1.0]", "D. [0.207, 0.085, 0.983, 0.985]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 997, "options": [ "A. extract -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)", "C. walk -> cook", "D. talk to (e.g., self, a person, a group) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1000, "options": [ "A. talk to (e.g., self, a person, a group) [0.019, 0.077, 0.921, 0.849] -> hug (a person) [0.064, 0.187, 0.973, 1.000] -> carry/hold (an object) [0.000, 0.311, 1.000, 1.000]", "B. smoke [0.000, 0.141, 0.888, 1.000] -> sail boat [0.000, 0.083, 1.000, 0.885] -> talk to (e.g., self, a person, a group) [0.086, 0.000, 1.000, 0.885]", "C. carry/hold (an object) [0.174, 0.139, 0.941, 0.959] -> swim [0.159, 0.186, 0.912, 0.984] -> smoke [0.191, 0.257, 0.841, 0.913]", "D. smoke [0.0, 0.113, 0.94, 0.99] -> talk to (e.g., self, a person, a group) [0.0, 0.113, 0.94, 0.99] -> carry/hold (an object) [0.0, 0.124, 0.975, 0.968]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.124, 0.975, 0.968] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1070, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.084, 0.103, 0.938, 0.983] location in the current frame performing the smoke?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1109, "options": [ "A. carry/hold (an object), get up, talk to (e.g., self, a person, a group)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, carry/hold (an object)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.097, 0.052, 0.59, 0.986] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1159, "options": [ "A. press", "B. walk", "C. hug (a person)", "D. smoke and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.212, 0.378, 0.272, 0.61] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1159, "options": [ "A. carry/hold (an object) [0.095, 0.259, 0.402, 0.448] -> carry/hold (an object) [0.266, 0.459, 0.352, 0.661] -> walk [0.017, 0.381, 0.384, 0.778]", "B. carry/hold (an object) [0.348, 0.200, 0.120, 0.598] -> watch (e.g., TV) [0.237, 0.231, 0.285, 0.703] -> talk to (e.g., self, a person, a group) [0.353, 0.237, 0.402, 0.691]", "C. carry/hold (an object) [0.035, 0.467, 0.398, 0.512] -> talk to (e.g., self, a person, a group) [0.276, 0.488, 0.230, 0.651] -> swim [0.125, 0.283, 0.290, 0.413]", "D. walk [0.212, 0.378, 0.272, 0.61] -> carry/hold (an object) [0.212, 0.378, 0.272, 0.61] -> talk to (e.g., self, a person, a group) [0.212, 0.378, 0.272, 0.61]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.212, 0.378, 0.272, 0.61] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1164, "options": [ "A. kick (a person) [0.107, 0.237, 0.487, 0.657] -> walk [0.249, 0.310, 0.482, 0.531] -> carry/hold (an object) [0.431, 0.137, 0.466, 0.649]", "B. lift (a person) [0.123, 0.346, 0.341, 0.623] -> talk to (e.g., self, a person, a group) [0.117, 0.201, 0.486, 0.665] -> walk [0.073, 0.263, 0.375, 0.574]", "C. walk [0.258, 0.281, 0.383, 0.646] -> carry/hold (an object) [0.258, 0.281, 0.383, 0.646] -> talk to (e.g., self, a person, a group) [0.258, 0.281, 0.383, 0.646]", "D. walk [0.067, 0.282, 0.426, 0.689] -> talk to (e.g., self, a person, a group) [0.330, 0.289, 0.561, 0.508] -> sail boat [0.068, 0.110, 0.411, 0.778]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.258, 0.281, 0.383, 0.646] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1168, "options": [ "A. walk [0.275, 0.142, 0.548, 0.991] -> carry/hold (an object) [0.275, 0.142, 0.548, 0.991] -> talk to (e.g., self, a person, a group) [0.275, 0.142, 0.548, 0.991]", "B. carry/hold (an object) [0.272, 0.267, 0.592, 1.000] -> lie/sleep [0.296, 0.183, 0.415, 1.000] -> walk [0.346, 0.091, 0.497, 0.997]", "C. carry/hold (an object) [0.212, 0.269, 0.528, 0.933] -> walk [0.097, 0.234, 0.356, 1.000] -> talk to (e.g., self, a person, a group) [0.380, 0.232, 0.582, 0.854]", "D. talk to (e.g., self, a person, a group) [0.194, 0.202, 0.582, 0.902] -> walk [0.102, 0.046, 0.414, 1.000] -> answer phone [0.108, 0.114, 0.460, 0.866]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.275, 0.142, 0.548, 0.991] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1181, "options": [ "A. [0.072, 0.069, 0.97, 0.991]", "B. [0.0, 0.0, 1.0, 0.741]", "C. [0.212, 0.378, 0.272, 0.61]", "D. [0.0, 0.344, 0.924, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1182, "options": [ "A. talk to (e.g., self, a person, a group) -> work on a computer", "B. carry/hold (an object) -> shoot", "C. jump/leap -> carry/hold (an object)", "D. talk to (e.g., self, a person, a group) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1183, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1381, "options": [ "A. [0.0, 0.328, 0.618, 1.0]", "B. [0.4, 0.0, 0.989, 1.0]", "C. [0.126, 0.1, 0.916, 0.995]", "D. [0.256, 0.309, 0.427, 0.651]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1385, "options": [ "A. [0.02, 0.146, 0.762, 0.998]", "B. [0.232, 0.166, 0.361, 0.676]", "C. [0.246, 0.342, 1.0, 0.947]", "D. [0.0, 0.446, 0.894, 1.0]" ], "question": "Where is the person currently performing the close (e.g., a door, a box) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1408, "options": [ "A. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "B. play musical instrument -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> ride (e.g., a bike, a car, a horse)", "D. talk to (e.g., self, a person, a group) -> hit (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1419, "options": [ "A. press [0.001, 0.287, 0.922, 0.799] -> talk to (e.g., self, a person, a group) [0.068, 0.028, 1.000, 1.000] -> carry/hold (an object) [0.000, 0.000, 0.904, 0.909]", "B. brush teeth [0.176, 0.245, 0.782, 1.000] -> smoke [0.000, 0.284, 0.988, 0.798] -> talk to (e.g., self, a person, a group) [0.000, 0.056, 1.000, 1.000]", "C. put down [0.184, 0.162, 0.794, 0.877] -> carry/hold (an object) [0.000, 0.271, 0.824, 1.000] -> smoke [0.113, 0.000, 0.865, 1.000]", "D. carry/hold (an object) [0.0, 0.111, 0.948, 0.986] -> smoke [0.0, 0.111, 0.948, 0.986] -> talk to (e.g., self, a person, a group) [0.0, 0.111, 0.948, 0.986]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.111, 0.948, 0.986] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1481, "options": [ "A. [0.506, 0.227, 0.872, 0.959]", "B. [0.282, 0.4, 0.61, 0.861]", "C. [0.235, 0.65, 0.387, 1.0]", "D. [0.467, 0.227, 0.625, 0.562]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1481, "options": [ "A. [0.426, 0.142, 0.669, 0.876]", "B. [0.291, 0.385, 0.607, 0.876]", "C. [0.204, 0.084, 0.412, 0.59]", "D. [0.431, 0.063, 0.746, 0.636]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1497, "options": [ "A. walk -> paint", "B. walk -> close (e.g., a door, a box)", "C. walk -> drive (e.g., a car, a truck)", "D. crawl -> close (e.g., a door, a box)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1530, "options": [ "A. [0.201, 0.222, 0.394, 0.621]", "B. [0.485, 0.0, 0.495, 0.523]", "C. [0.459, 0.192, 0.674, 0.823]", "D. [0.061, 0.0, 0.171, 0.407]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1543, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1785, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "B. sit, listen to (a person)", "C. sit, talk to (e.g., self, a person, a group)", "D. sit, carry/hold (an object)" ], "question": "What action is the person at the location [0.269, 0.357, 0.397, 0.645] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1794, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1794 ]

[ { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 914, "options": [ "A. [0.661, 0.404, 0.688, 1.0]", "B. [0.793, 0.317, 0.981, 0.929]", "C. [0.273, 0.866, 0.36, 0.993]", "D. [0.911, 0.072, 1.0, 0.748]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1033, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.168, 0.193, 0.846, 0.996] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1056, "options": [ "A. [0.0, 0.261, 0.495, 1.0]", "B. [0.26, 0.089, 0.852, 0.994]", "C. [0.0, 0.0, 0.417, 0.715]", "D. [0.047, 0.036, 0.694, 0.954]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1187, "options": [ "A. stand, hand wave, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, listen to (a person), talk to (e.g., self, a person, a group)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.007, 0.015, 0.449, 0.988] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1192, "options": [ "A. 5", "B. 7", "C. 6", "D. 3" ], "question": "How many people were performing the carry/hold (an object) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1212, "options": [ "A. walk, watch (a person)", "B. walk, carry/hold (an object), watch (a person)", "C. sit, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.802, 0.397, 0.999, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1239, "options": [ "A. talk to (e.g., self, a person, a group) [0.465, 0.411, 0.518, 0.784] -> dance [0.527, 0.232, 0.763, 0.927] -> walk [0.514, 0.106, 0.632, 0.949]", "B. take (an object) from (a person) [0.338, 0.293, 0.709, 0.877] -> walk [0.357, 0.048, 0.746, 1.000] -> carry/hold (an object) [0.524, 0.364, 0.888, 0.691]", "C. walk [0.471, 0.225, 0.697, 0.81] -> carry/hold (an object) [0.471, 0.225, 0.697, 0.81] -> talk to (e.g., self, a person, a group) [0.471, 0.225, 0.697, 0.81]", "D. open (e.g., a window, a car door) [0.314, 0.342, 0.642, 0.941] -> carry/hold (an object) [0.490, 0.090, 0.565, 1.000] -> walk [0.445, 0.202, 0.699, 0.932]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.471, 0.225, 0.697, 0.81] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1337, "options": [ "A. walk [0.164, 0.198, 0.425, 0.774] -> talk to (e.g., self, a person, a group) [0.252, 0.061, 0.357, 0.981] -> walk [0.330, 0.127, 0.458, 0.963]", "B. carry/hold (an object) [0.374, 0.233, 0.461, 0.982] -> talk to (e.g., self, a person, a group) [0.396, 0.066, 0.558, 1.000] -> crawl [0.015, 0.000, 0.430, 1.000]", "C. walk [0.2, 0.149, 0.415, 0.959] -> carry/hold (an object) [0.2, 0.149, 0.415, 0.959] -> talk to (e.g., self, a person, a group) [0.2, 0.149, 0.415, 0.959]", "D. walk [0.148, 0.000, 0.303, 0.834] -> dig [0.226, 0.049, 0.254, 1.000] -> talk to (e.g., self, a person, a group) [0.146, 0.312, 0.336, 0.802]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.2, 0.149, 0.415, 0.959] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1338, "options": [ "A. walk [0.2, 0.149, 0.415, 0.959] -> carry/hold (an object) [0.2, 0.149, 0.415, 0.959] -> talk to (e.g., self, a person, a group) [0.2, 0.149, 0.415, 0.959]", "B. walk [0.194, 0.236, 0.279, 0.906] -> talk to (e.g., self, a person, a group) [0.366, 0.267, 0.593, 1.000] -> run/jog [0.050, 0.156, 0.312, 1.000]", "C. walk [0.148, 0.275, 0.483, 0.902] -> carry/hold (an object) [0.091, 0.298, 0.344, 0.952] -> text on/look at a cellphone [0.126, 0.224, 0.297, 0.978]", "D. walk [0.307, 0.302, 0.246, 1.000] -> talk to (e.g., self, a person, a group) [0.261, 0.241, 0.433, 0.826] -> martial art [0.228, 0.000, 0.344, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.2, 0.149, 0.415, 0.959] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1408, "options": [ "A. [0.002, 0.177, 0.835, 1.0]", "B. [0.0, 0.0, 1.0, 0.713]", "C. [0.418, 0.004, 1.0, 0.911]", "D. [0.175, 0.059, 0.968, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1414, "options": [ "A. work on a computer", "B. bend/bow (at the waist) and carry/hold (an object)", "C. eat", "D. walk" ], "question": "What action is the person currently in the [0.743, 0.429, 0.895, 0.936] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1500, "options": [ "A. 6", "B. 5", "C. 8", "D. 3" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1543, "options": [ "A. 3", "B. 5", "C. 2", "D. 4" ], "question": "How many people in the current frame are performing the action: lie/sleep?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1566, "options": [ "A. walk [0.232, 0.097, 0.413, 0.55] -> carry/hold (an object) [0.232, 0.097, 0.413, 0.55] -> talk to (e.g., self, a person, a group) [0.232, 0.097, 0.413, 0.55]", "B. walk [0.280, 0.194, 0.239, 0.359] -> swim [0.098, 0.009, 0.340, 0.725] -> carry/hold (an object) [0.105, 0.279, 0.237, 0.743]", "C. lift/pick up [0.124, 0.150, 0.318, 0.637] -> talk to (e.g., self, a person, a group) [0.413, 0.166, 0.444, 0.528] -> carry/hold (an object) [0.428, 0.043, 0.309, 0.603]", "D. walk [0.148, 0.002, 0.328, 0.479] -> talk to (e.g., self, a person, a group) [0.327, 0.230, 0.434, 0.491] -> play with kids [0.387, 0.141, 0.531, 0.378]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.232, 0.097, 0.413, 0.55] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1656, "options": [ "A. get up [0.665, 0.423, 0.615, 0.575] -> talk to (e.g., self, a person, a group) [0.632, 0.452, 0.569, 0.741] -> carry/hold (an object) [0.667, 0.358, 0.636, 0.850]", "B. play board game [0.625, 0.246, 0.747, 0.764] -> talk to (e.g., self, a person, a group) [0.613, 0.578, 0.794, 0.744] -> carry/hold (an object) [0.577, 0.370, 0.926, 0.714]", "C. carry/hold (an object) [0.632, 0.476, 0.844, 0.533] -> walk [0.630, 0.574, 0.596, 0.640] -> walk [0.858, 0.408, 0.813, 0.701]", "D. walk [0.66, 0.384, 0.752, 0.688] -> carry/hold (an object) [0.66, 0.384, 0.752, 0.688] -> talk to (e.g., self, a person, a group) [0.66, 0.384, 0.752, 0.688]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.66, 0.384, 0.752, 0.688] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1670, "options": [ "A. walk, carry/hold (an object)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. stand, watch (a person)", "D. stand, carry/hold (an object), watch (a person)" ], "question": "What action is the person at the location [0.0, 0.187, 0.168, 0.861] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1674, "options": [ "A. 3", "B. 6", "C. 4", "D. 5" ], "question": "How many people were performing the listen to (a person) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1674 ]

[ { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 953, "options": [ "A. talk to (e.g., self, a person, a group) [0.000, 0.024, 0.439, 0.959] -> play with kids [0.000, 0.000, 0.210, 0.880] -> carry/hold (an object) [0.187, 0.000, 0.371, 0.719] -> walk [0.000, 0.000, 0.204, 0.835]", "B. talk to (e.g., self, a person, a group) [0.014, 0.000, 0.218, 0.757] -> carry/hold (an object) [0.210, 0.000, 0.500, 0.803] -> walk [0.000, 0.142, 0.146, 0.707] -> dance [0.099, 0.082, 0.143, 0.866]", "C. walk [0.012, 0.023, 0.328, 0.891] -> carry/hold (an object) [0.012, 0.023, 0.328, 0.891] -> smoke [0.012, 0.023, 0.328, 0.891] -> talk to (e.g., self, a person, a group) [0.012, 0.023, 0.328, 0.891]", "D. talk to (e.g., self, a person, a group) [0.000, 0.000, 0.393, 0.838] -> turn (e.g., a screwdriver) [0.197, 0.099, 0.413, 1.000] -> carry/hold (an object) [0.000, 0.080, 0.138, 0.881] -> smoke [0.000, 0.112, 0.331, 0.991]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.012, 0.023, 0.328, 0.891] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 961, "options": [ "A. drive (e.g., a car, a truck) [0.033, 0.107, 0.335, 0.808] -> carry/hold (an object) [0.023, 0.010, 0.373, 1.000] -> talk to (e.g., self, a person, a group) [0.164, 0.147, 0.420, 1.000] -> walk [0.000, 0.094, 0.216, 1.000]", "B. talk to (e.g., self, a person, a group) [0.000, 0.084, 0.560, 0.861] -> carry/hold (an object) [0.027, 0.000, 0.269, 0.917] -> walk [0.177, 0.000, 0.401, 0.931] -> chop [0.000, 0.208, 0.541, 1.000]", "C. walk [0.0, 0.038, 0.375, 0.993] -> carry/hold (an object) [0.0, 0.038, 0.375, 0.993] -> smoke [0.0, 0.038, 0.375, 0.993] -> talk to (e.g., self, a person, a group) [0.0, 0.038, 0.375, 0.993]", "D. walk [0.141, 0.000, 0.184, 0.999] -> carry/hold (an object) [0.000, 0.018, 0.348, 1.000] -> fall down [0.007, 0.086, 0.271, 0.852] -> smoke [0.000, 0.219, 0.339, 0.906]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.038, 0.375, 0.993] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 962, "options": [ "A. [0.252, 0.179, 0.502, 1.0]", "B. [0.0, 0.078, 0.288, 0.983]", "C. [0.434, 0.151, 0.725, 0.973]", "D. [0.033, 0.036, 0.368, 0.719]" ], "question": "Where was the person currently performing the smoke in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1013, "options": [ "A. [0.411, 0.321, 0.998, 0.99]", "B. [0.029, 0.0, 0.818, 0.772]", "C. [0.01, 0.271, 0.994, 1.0]", "D. [0.168, 0.005, 1.0, 0.967]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1051, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1086, "options": [ "A. 3", "B. 2", "C. 7", "D. 4" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1136, "options": [ "A. 7", "B. 5", "C. 4", "D. 3" ], "question": "How many people were performing the listen to (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1161, "options": [ "A. walk [0.038, 0.112, 0.189, 0.615] -> carry/hold (an object) [0.038, 0.112, 0.189, 0.615] -> talk to (e.g., self, a person, a group) [0.038, 0.112, 0.189, 0.615]", "B. play musical instrument [0.000, 0.163, 0.191, 0.713] -> talk to (e.g., self, a person, a group) [0.000, 0.227, 0.134, 0.625] -> walk [0.000, 0.013, 0.162, 0.705]", "C. carry/hold (an object) [0.000, 0.293, 0.324, 0.533] -> fight/hit (a person) [0.143, 0.210, 0.218, 0.795] -> walk [0.170, 0.069, 0.185, 0.524]", "D. crawl [0.000, 0.000, 0.142, 0.681] -> carry/hold (an object) [0.000, 0.000, 0.093, 0.528] -> talk to (e.g., self, a person, a group) [0.000, 0.258, 0.271, 0.732]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.038, 0.112, 0.189, 0.615] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1183, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1232, "options": [ "A. 4", "B. 6", "C. 5", "D. 3" ], "question": "How many people were performing the walk in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1303, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, watch (a person)", "C. sit, ride (e.g., a bike, a car, a horse)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.059, 0.222, 0.116, 0.362] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1322, "options": [ "A. walk [0.000, 0.160, 0.271, 1.000] -> answer phone [0.000, 0.130, 0.155, 0.954] -> talk to (e.g., self, a person, a group) [0.050, 0.175, 0.402, 0.798]", "B. walk [0.000, 0.118, 0.319, 0.806] -> watch (e.g., TV) [0.000, 0.129, 0.240, 1.000] -> carry/hold (an object) [0.163, 0.337, 0.095, 1.000]", "C. walk [0.006, 0.227, 0.282, 0.983] -> carry/hold (an object) [0.006, 0.227, 0.282, 0.983] -> talk to (e.g., self, a person, a group) [0.006, 0.227, 0.282, 0.983]", "D. dress/put on clothing [0.172, 0.070, 0.436, 0.947] -> carry/hold (an object) [0.020, 0.406, 0.230, 0.949] -> walk [0.105, 0.245, 0.169, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.006, 0.227, 0.282, 0.983] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1325, "options": [ "A. [0.212, 0.0, 0.637, 0.703]", "B. [0.331, 0.065, 0.793, 1.0]", "C. [0.105, 0.33, 0.605, 1.0]", "D. [0.726, 0.166, 0.879, 0.925]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1403, "options": [ "A. walk -> talk to (e.g., self, a person, a group) -> carry/hold (an object) -> swim", "B. talk to (e.g., self, a person, a group) -> crawl -> walk -> close (e.g., a door, a box)", "C. walk -> carry/hold (an object) -> talk to (e.g., self, a person, a group) -> crawl", "D. sing to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group) -> carry/hold (an object) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1426, "options": [ "A. 7", "B. 3", "C. 2", "D. 4" ], "question": "How many people in the current frame are performing the action: run/jog?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1429, "options": [ "A. talk to (e.g., self, a person, a group) -> push (an object)", "B. run/jog -> talk to (e.g., self, a person, a group)", "C. lie/sleep -> run/jog", "D. push (another person) -> run/jog" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1526, "options": [ "A. talk to (e.g., self, a person, a group)", "B. play with pets", "C. carry/hold (an object)", "D. eat" ], "question": "What action is the person currently in the [0.021, 0.085, 0.705, 0.978] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1544, "options": [ "A. press -> carry/hold (an object)", "B. swim -> carry/hold (an object)", "C. bend/bow (at the waist) -> turn (e.g., a screwdriver)", "D. bend/bow (at the waist) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1557, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.01, 0.677, 0.984] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1562, "options": [ "A. kick (an object) -> hit (an object) -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> hit (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> hit (an object) -> fight/hit (a person)", "D. enter -> carry/hold (an object) -> hit (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1641, "options": [ "A. talk to (e.g., self, a person, a group) [0.358, 0.080, 0.630, 0.798] -> carry/hold (an object) [0.618, 0.371, 0.671, 1.000] -> take (an object) from (a person) [0.456, 0.140, 0.862, 1.000]", "B. drink [0.316, 0.331, 0.660, 0.811] -> talk to (e.g., self, a person, a group) [0.282, 0.369, 0.768, 0.870] -> carry/hold (an object) [0.633, 0.260, 0.688, 1.000]", "C. carry/hold (an object) [0.451, 0.185, 0.682, 0.989] -> grab (a person) [0.451, 0.185, 0.682, 0.989] -> talk to (e.g., self, a person, a group) [0.451, 0.185, 0.682, 0.989]", "D. carry/hold (an object) [0.580, 0.276, 0.691, 0.912] -> drink [0.261, 0.252, 0.672, 1.000] -> talk to (e.g., self, a person, a group) [0.608, 0.205, 0.517, 0.973]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.451, 0.185, 0.682, 0.989] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1649, "options": [ "A. [0.203, 0.0, 0.632, 0.877]", "B. [0.451, 0.185, 0.682, 0.989]", "C. [0.037, 0.391, 0.798, 0.988]", "D. [0.705, 0.102, 0.876, 0.741]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1651, "options": [ "A. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "B. hand wave -> carry/hold (an object)", "C. hug (a person) -> carry/hold (an object)", "D. carry/hold (an object) -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1778, "options": [ "A. sit, watch (a person)", "B. sit, carry/hold (an object), watch (a person)", "C. stand, carry/hold (an object)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.004, 0.096, 0.42, 0.993] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 900, 1778 ]

[ { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 986, "options": [ "A. sit, answer phone", "B. stand, answer phone", "C. sit, talk to (e.g., self, a person, a group)", "D. dance, watch (a person)" ], "question": "What action is the person at the location [0.017, 0.051, 0.621, 0.969] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1037, "options": [ "A. [0.546, 0.033, 0.996, 0.968]", "B. [0.3, 0.093, 1.0, 0.688]", "C. [0.636, 0.0, 0.989, 0.678]", "D. [0.529, 0.296, 0.71, 0.915]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1069, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1119, "options": [ "A. [0.051, 0.101, 0.343, 0.976]", "B. [0.032, 0.061, 0.876, 1.0]", "C. [0.007, 0.294, 0.638, 1.0]", "D. [0.0, 0.02, 0.131, 0.695]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1211, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1231, "options": [ "A. [0.0, 0.323, 0.753, 0.967]", "B. [0.136, 0.081, 1.0, 1.0]", "C. [0.567, 0.109, 0.969, 0.994]", "D. [0.112, 0.034, 1.0, 0.741]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1294, "options": [ "A. carry/hold (an object) [0.556, 0.035, 0.996, 0.916] -> talk to (e.g., self, a person, a group) [0.556, 0.035, 0.996, 0.916] -> bend/bow (at the waist) [0.448, 0.0, 0.993, 0.929]", "B. crawl [0.623, 0.000, 0.926, 1.000] -> talk to (e.g., self, a person, a group) [0.678, 0.166, 0.838, 1.000] -> bend/bow (at the waist) [0.451, 0.150, 1.000, 1.000]", "C. bend/bow (at the waist) [0.339, 0.006, 0.882, 0.786] -> carry/hold (an object) [0.658, 0.000, 1.000, 0.771] -> row boat [0.754, 0.000, 1.000, 1.000]", "D. talk to (e.g., self, a person, a group) [0.699, 0.000, 1.000, 0.822] -> carry/hold (an object) [0.705, 0.109, 0.853, 0.745] -> hug (a person) [0.501, 0.000, 0.834, 0.914]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.448, 0.0, 0.993, 0.929] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1301, "options": [ "A. talk to (e.g., self, a person, a group) [0.000, 0.304, 0.463, 1.000] -> read [0.043, 0.045, 0.399, 1.000] -> sing to (e.g., self, a person, a group) [0.000, 0.097, 0.655, 0.898]", "B. dress/put on clothing [0.122, 0.230, 0.544, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.039, 0.411, 0.880] -> read [0.167, 0.293, 0.594, 0.817]", "C. carry/hold (an object) [0.022, 0.156, 0.456, 0.921] -> read [0.022, 0.156, 0.456, 0.921] -> talk to (e.g., self, a person, a group) [0.022, 0.156, 0.456, 0.921]", "D. talk to (e.g., self, a person, a group) [0.108, 0.243, 0.439, 0.835] -> read [0.173, 0.335, 0.591, 0.936] -> dance [0.040, 0.070, 0.443, 0.957]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.022, 0.156, 0.456, 0.921] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1307, "options": [ "A. [0.588, 0.031, 0.991, 0.977]", "B. [0.0, 0.413, 0.232, 0.862]", "C. [0.027, 0.147, 0.463, 0.934]", "D. [0.247, 0.0, 0.71, 0.994]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1309, "options": [ "A. walk [0.139, 0.192, 0.567, 0.764] -> read [0.079, 0.306, 0.359, 1.000] -> talk to (e.g., self, a person, a group) [0.095, 0.056, 0.421, 0.779]", "B. carry/hold (an object) [0.000, 0.025, 0.506, 1.000] -> give/serve (an object) to (a person) [0.080, 0.290, 0.242, 0.752] -> read [0.169, 0.000, 0.468, 0.754]", "C. carry/hold (an object) [0.016, 0.146, 0.439, 0.929] -> read [0.016, 0.146, 0.439, 0.929] -> talk to (e.g., self, a person, a group) [0.016, 0.146, 0.439, 0.929]", "D. carry/hold (an object) [0.069, 0.010, 0.526, 0.796] -> push (an object) [0.000, 0.151, 0.490, 0.992] -> read [0.164, 0.081, 0.576, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.016, 0.146, 0.439, 0.929] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1373, "options": [ "A. sit, listen to (a person), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. stand, answer phone", "D. sit, answer phone" ], "question": "What action is the person at the location [0.076, 0.197, 0.361, 0.981] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1398, "options": [ "A. close (e.g., a door, a box) [0.536, 0.239, 0.961, 0.989] -> talk to (e.g., self, a person, a group) [0.828, 0.159, 0.901, 1.000] -> read [0.480, 0.413, 1.000, 0.791]", "B. hand wave [0.750, 0.348, 0.812, 1.000] -> read [0.598, 0.134, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.723, 0.090, 0.885, 0.710]", "C. carry/hold (an object) [0.63, 0.285, 1.0, 0.852] -> read [0.63, 0.285, 1.0, 0.852] -> talk to (e.g., self, a person, a group) [0.63, 0.285, 1.0, 0.852]", "D. kiss (a person) [0.735, 0.191, 1.000, 0.830] -> carry/hold (an object) [0.700, 0.311, 0.982, 0.904] -> talk to (e.g., self, a person, a group) [0.436, 0.371, 1.000, 0.851]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.63, 0.285, 1.0, 0.852] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1419, "options": [ "A. [0.648, 0.285, 1.0, 0.887]", "B. [0.002, 0.229, 0.359, 1.0]", "C. [0.355, 0.163, 1.0, 0.651]", "D. [0.764, 0.577, 1.0, 1.0]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1426, "options": [ "A. carry/hold (an object) [0.65, 0.291, 0.987, 0.797] -> read [0.65, 0.291, 0.987, 0.797] -> talk to (e.g., self, a person, a group) [0.65, 0.291, 0.987, 0.797]", "B. read [0.564, 0.374, 1.000, 0.680] -> drive (e.g., a car, a truck) [0.478, 0.468, 1.000, 0.796] -> talk to (e.g., self, a person, a group) [0.728, 0.129, 0.805, 0.670]", "C. carry/hold (an object) [0.523, 0.104, 0.948, 0.768] -> hit (an object) [0.668, 0.164, 0.925, 0.996] -> talk to (e.g., self, a person, a group) [0.477, 0.212, 1.000, 0.713]", "D. carry/hold (an object) [0.806, 0.387, 0.886, 0.928] -> push (another person) [0.816, 0.278, 1.000, 0.893] -> talk to (e.g., self, a person, a group) [0.767, 0.092, 1.000, 0.987]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.65, 0.291, 0.987, 0.797] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1432, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.135, 0.194, 0.519, 0.987] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1556, "options": [ "A. [0.49, 0.358, 0.955, 0.535]", "B. [0.617, 0.59, 0.79, 0.732]", "C. [0.477, 0.349, 0.661, 0.801]", "D. [0.01, 0.148, 0.764, 0.963]" ], "question": "Where is the person currently performing the answer phone located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1575, "options": [ "A. brush teeth", "B. fight/hit (a person)", "C. talk to (e.g., self, a person, a group)", "D. bend/bow (at the waist)" ], "question": "What action is the person currently in the [0.487, 0.089, 0.871, 0.932] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1623, "options": [ "A. [0.256, 0.187, 0.751, 0.907]", "B. [0.028, 0.155, 0.48, 0.977]", "C. [0.383, 0.497, 0.566, 0.993]", "D. [0.195, 0.014, 0.56, 0.713]" ], "question": "Where was the person currently performing the grab (a person) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1706, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1795, "options": [ "A. [0.011, 0.095, 0.727, 0.979]", "B. [0.215, 0.387, 1.0, 1.0]", "C. [0.0, 0.0, 0.943, 0.772]", "D. [0.507, 0.089, 0.889, 0.946]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1795 ]

[ { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 915, "options": [ "A. stand, listen to (a person)", "B. stand, carry/hold (an object)", "C. sit, watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.136, 0.124, 0.262, 0.816] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 921, "options": [ "A. fight/hit (a person) -> carry/hold (an object)", "B. carry/hold (an object) -> crouch/kneel", "C. carry/hold (an object) -> drink", "D. carry/hold (an object) -> ride (e.g., a bike, a car, a horse)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 927, "options": [ "A. climb (e.g., a mountain) -> carry/hold (an object)", "B. carry/hold (an object) -> crouch/kneel", "C. crouch/kneel -> enter", "D. crouch/kneel -> press" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 980, "options": [ "A. walk [0.511, 0.288, 0.642, 0.968] -> carry/hold (an object) [0.511, 0.288, 0.642, 0.968] -> talk to (e.g., self, a person, a group) [0.511, 0.288, 0.642, 0.968]", "B. carry/hold (an object) [0.353, 0.136, 0.693, 0.845] -> walk [0.668, 0.193, 0.601, 0.966] -> talk to (e.g., self, a person, a group) [0.321, 0.396, 0.562, 1.000]", "C. talk to (e.g., self, a person, a group) [0.625, 0.108, 0.512, 0.781] -> walk [0.387, 0.325, 0.499, 1.000] -> answer phone [0.433, 0.340, 0.664, 0.795]", "D. walk [0.696, 0.347, 0.495, 0.850] -> talk to (e.g., self, a person, a group) [0.352, 0.391, 0.646, 1.000] -> jump/leap [0.699, 0.200, 0.735, 0.911]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.511, 0.288, 0.642, 0.968] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1002, "options": [ "A. drink -> write", "B. dig -> drink", "C. drink -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> hand clap" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1067, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.004, 0.139, 0.308, 0.995] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1122, "options": [ "A. carry/hold (an object) [0.396, 0.096, 0.866, 0.984] -> give/serve (an object) to (a person) [0.396, 0.096, 0.866, 0.984] -> talk to (e.g., self, a person, a group) [0.396, 0.096, 0.866, 0.984]", "B. give/serve (an object) to (a person) [0.471, 0.287, 1.000, 1.000] -> carry/hold (an object) [0.252, 0.005, 1.000, 1.000] -> give/serve (an object) to (a person) [0.503, 0.000, 0.812, 1.000]", "C. swim [0.454, 0.081, 0.769, 1.000] -> carry/hold (an object) [0.589, 0.000, 0.801, 1.000] -> talk to (e.g., self, a person, a group) [0.230, 0.147, 1.000, 0.835]", "D. give/serve (an object) to (a person) [0.309, 0.076, 1.000, 0.946] -> talk to (e.g., self, a person, a group) [0.587, 0.046, 0.742, 1.000] -> climb (e.g., a mountain) [0.450, 0.276, 0.937, 0.981]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.396, 0.096, 0.866, 0.984] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1123, "options": [ "A. lie/sleep", "B. clink glass", "C. drink", "D. talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.037, 0.597, 0.226, 0.985] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1139, "options": [ "A. [0.029, 0.601, 0.237, 0.984]", "B. [0.029, 0.303, 0.268, 0.828]", "C. [0.003, 0.0, 0.847, 0.986]", "D. [0.154, 0.748, 0.508, 0.928]" ], "question": "Where was the person currently performing the lie/sleep in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1182, "options": [ "A. [0.207, 0.0, 1.0, 0.799]", "B. [0.323, 0.29, 1.0, 0.984]", "C. [0.287, 0.201, 0.661, 0.913]", "D. [0.482, 0.0, 1.0, 0.972]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1216, "options": [ "A. sit, watch (a person)", "B. sit, carry/hold (an object), listen to (a person)", "C. sit, talk to (e.g., self, a person, a group), watch (a person)", "D. walk, lift/pick up" ], "question": "What action is the person at the location [0.24, 0.141, 0.457, 0.829] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1383, "options": [ "A. 7", "B. 4", "C. 3", "D. 5" ], "question": "How many people were performing the watch (a person) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1398, "options": [ "A. run/jog [0.503, 0.342, 0.435, 1.000] -> walk [0.429, 0.423, 0.636, 1.000] -> lift/pick up [0.267, 0.188, 0.553, 1.000]", "B. walk [0.34, 0.298, 0.503, 0.984] -> carry/hold (an object) [0.34, 0.298, 0.503, 0.984] -> run/jog [0.319, 0.223, 0.561, 0.979]", "C. run/jog [0.313, 0.340, 0.682, 1.000] -> carry/hold (an object) [0.406, 0.356, 0.392, 1.000] -> stir [0.260, 0.101, 0.412, 0.824]", "D. walk [0.279, 0.237, 0.388, 0.956] -> play board game [0.155, 0.286, 0.388, 0.979] -> run/jog [0.508, 0.204, 0.466, 0.995]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.319, 0.223, 0.561, 0.979] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1398, "options": [ "A. walk -> clink glass -> run/jog", "B. walk -> cook -> carry/hold (an object)", "C. walk -> carry/hold (an object) -> run/jog", "D. run/jog -> walk -> smoke" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1488, "options": [ "A. get up [0.270, 0.105, 0.825, 0.808] -> carry/hold (an object) [0.226, 0.038, 0.807, 0.804] -> shoot [0.010, 0.324, 0.576, 1.000]", "B. carry/hold (an object) [0.324, 0.205, 0.838, 0.900] -> shoot [0.057, 0.007, 0.844, 1.000] -> bend/bow (at the waist) [0.182, 0.158, 0.498, 1.000]", "C. shoot [0.149, 0.146, 0.648, 0.786] -> carry/hold (an object) [0.253, 0.000, 0.812, 1.000] -> close (e.g., a door, a box) [0.000, 0.270, 0.664, 0.814]", "D. bend/bow (at the waist) [0.148, 0.135, 0.656, 0.946] -> carry/hold (an object) [0.148, 0.135, 0.656, 0.946] -> shoot [0.148, 0.135, 0.656, 0.946]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.148, 0.135, 0.656, 0.946] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1499, "options": [ "A. crouch/kneel -> shovel -> bend/bow (at the waist)", "B. carry/hold (an object) -> bend/bow (at the waist) -> shovel", "C. bend/bow (at the waist) -> carry/hold (an object) -> crouch/kneel", "D. bend/bow (at the waist) -> martial art -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1514, "options": [ "A. 3", "B. 7", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: bend/bow (at the waist)?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1629, "options": [ "A. stand, carry/hold (an object)", "B. sit, watch (a person)", "C. sit, listen to (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.018, 0.128, 0.512, 0.983] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1639, "options": [ "A. [0.168, 0.253, 0.396, 0.983]", "B. [0.592, 0.0, 1.0, 0.703]", "C. [0.464, 0.0, 0.737, 0.696]", "D. [0.496, 0.024, 1.0, 0.992]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1671, "options": [ "A. Yes", "B. No" ], "question": "Are there still 6 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1715, "options": [ "A. 3", "B. 6", "C. 4", "D. 5" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1764, "options": [ "A. walk [0.271, 0.313, 0.516, 0.801] -> grab (a person) [0.362, 0.282, 0.410, 0.809] -> carry/hold (an object) [0.089, 0.329, 0.480, 1.000]", "B. walk [0.144, 0.000, 0.553, 1.000] -> push (an object) [0.168, 0.236, 0.360, 1.000] -> sail boat [0.293, 0.148, 0.561, 1.000]", "C. walk [0.284, 0.159, 0.551, 0.99] -> carry/hold (an object) [0.284, 0.159, 0.551, 0.99] -> push (an object) [0.284, 0.159, 0.551, 0.99]", "D. push (an object) [0.299, 0.000, 0.451, 0.986] -> exit [0.089, 0.136, 0.628, 0.813] -> carry/hold (an object) [0.090, 0.303, 0.725, 0.841]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.284, 0.159, 0.551, 0.99] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1768, "options": [ "A. [0.43, 0.606, 0.505, 1.0]", "B. [0.419, 0.347, 1.0, 0.609]", "C. [0.055, 0.006, 0.611, 0.962]", "D. [0.665, 0.487, 0.752, 0.87]" ], "question": "Where is the person currently performing the ride (e.g., a bike, a car, a horse) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1768 ]

[ { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1022, "options": [ "A. walk -> crouch/kneel", "B. bend/bow (at the waist) -> walk", "C. pull (an object) -> walk", "D. walk -> hand wave" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1031, "options": [ "A. crouch/kneel -> eat", "B. carry/hold (an object) -> crouch/kneel", "C. carry/hold (an object) -> play musical instrument", "D. crouch/kneel -> lift/pick up" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1041, "options": [ "A. [0.372, 0.283, 0.73, 0.989]", "B. [0.559, 0.359, 0.985, 1.0]", "C. [0.584, 0.093, 1.0, 0.792]", "D. [0.473, 0.613, 0.853, 0.956]" ], "question": "Where is the person currently performing the run/jog located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1055, "options": [ "A. talk to (e.g., self, a person, a group) [0.409, 0.426, 0.632, 0.868] -> crouch/kneel [0.424, 0.334, 0.615, 0.863] -> grab (a person) [0.424, 0.334, 0.615, 0.863]", "B. grab (a person) [0.591, 0.211, 0.524, 0.951] -> crouch/kneel [0.433, 0.288, 0.700, 0.928] -> dig [0.345, 0.275, 0.706, 1.000]", "C. crouch/kneel [0.294, 0.342, 0.723, 0.946] -> open (e.g., a window, a car door) [0.525, 0.360, 0.667, 0.858] -> talk to (e.g., self, a person, a group) [0.550, 0.420, 0.632, 1.000]", "D. grab (a person) [0.541, 0.362, 0.809, 1.000] -> extract [0.302, 0.487, 0.536, 0.759] -> crouch/kneel [0.551, 0.172, 0.721, 0.982]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.424, 0.334, 0.615, 0.863] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1107, "options": [ "A. [0.331, 0.029, 0.704, 0.774]", "B. [0.0, 0.455, 0.489, 1.0]", "C. [0.214, 0.17, 0.57, 0.98]", "D. [0.559, 0.359, 0.985, 1.0]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1114, "options": [ "A. walk -> carry/hold (an object) -> shoot", "B. carry/hold (an object) -> walk -> watch (e.g., TV)", "C. carry/hold (an object) -> ride (e.g., a bike, a car, a horse) -> shoot", "D. carry/hold (an object) -> shoot -> dress/put on clothing" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1145, "options": [ "A. [0.259, 0.184, 0.607, 0.995]", "B. [0.436, 0.0, 0.517, 0.779]", "C. [0.118, 0.203, 0.32, 0.812]", "D. [0.332, 0.065, 0.938, 0.996]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1147, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.32, 0.041, 0.592, 0.961] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1161, "options": [ "A. walk -> talk to (e.g., self, a person, a group)", "B. shoot -> walk", "C. walk -> hand shake", "D. kiss (a person) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1173, "options": [ "A. [0.485, 0.0, 1.0, 0.702]", "B. [0.895, 0.272, 0.972, 0.639]", "C. [0.723, 0.377, 1.0, 1.0]", "D. [0.518, 0.089, 0.979, 1.0]" ], "question": "What location in the frame is the person currently in the [0.242, 0.17, 0.624, 0.994] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1226, "options": [ "A. [0.375, 0.103, 0.993, 0.995]", "B. [0.12, 0.065, 0.471, 0.776]", "C. [0.318, 0.186, 0.577, 0.991]", "D. [0.574, 0.459, 0.743, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1306, "options": [ "A. take a photo -> put down", "B. drink -> put down", "C. carry/hold (an object) -> put down", "D. kick (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1366, "options": [ "A. carry/hold (an object) [0.267, 0.178, 0.597, 0.917] -> kick (an object) [0.328, 0.034, 0.687, 0.677] -> talk to (e.g., self, a person, a group) [0.351, 0.341, 0.517, 0.680]", "B. walk [0.278, 0.171, 0.703, 0.852] -> talk to (e.g., self, a person, a group) [0.404, 0.263, 0.401, 0.948] -> hug (a person) [0.471, 0.102, 0.401, 0.982]", "C. exit [0.504, 0.008, 0.692, 0.786] -> talk to (e.g., self, a person, a group) [0.438, 0.148, 0.625, 0.649] -> walk [0.477, 0.215, 0.378, 0.845]", "D. walk [0.42, 0.194, 0.543, 0.794] -> carry/hold (an object) [0.42, 0.194, 0.543, 0.794] -> talk to (e.g., self, a person, a group) [0.42, 0.194, 0.543, 0.794]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.42, 0.194, 0.543, 0.794] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1374, "options": [ "A. walk [0.514, 0.131, 0.691, 0.937] -> carry/hold (an object) [0.514, 0.131, 0.691, 0.937] -> talk to (e.g., self, a person, a group) [0.514, 0.131, 0.691, 0.937]", "B. talk to (e.g., self, a person, a group) [0.598, 0.003, 0.509, 1.000] -> walk [0.656, 0.003, 0.792, 1.000] -> fishing [0.707, 0.090, 0.492, 0.941]", "C. take (an object) from (a person) [0.626, 0.255, 0.710, 1.000] -> walk [0.687, 0.004, 0.707, 1.000] -> talk to (e.g., self, a person, a group) [0.608, 0.135, 0.678, 0.805]", "D. brush teeth [0.696, 0.248, 0.794, 1.000] -> walk [0.400, 0.072, 0.690, 0.942] -> talk to (e.g., self, a person, a group) [0.447, 0.167, 0.577, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.514, 0.131, 0.691, 0.937] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1379, "options": [ "A. walk [0.38, 0.121, 0.566, 0.995] -> carry/hold (an object) [0.38, 0.121, 0.566, 0.995] -> talk to (e.g., self, a person, a group) [0.38, 0.121, 0.566, 0.995]", "B. carry/hold (an object) [0.534, 0.045, 0.449, 1.000] -> take a photo [0.502, 0.039, 0.633, 1.000] -> talk to (e.g., self, a person, a group) [0.289, 0.232, 0.538, 1.000]", "C. talk to (e.g., self, a person, a group) [0.384, 0.265, 0.390, 0.945] -> walk [0.428, 0.202, 0.439, 0.975] -> get up [0.326, 0.025, 0.737, 0.842]", "D. give/serve (an object) to (a person) [0.325, 0.259, 0.530, 1.000] -> walk [0.278, 0.124, 0.763, 1.000] -> talk to (e.g., self, a person, a group) [0.364, 0.209, 0.576, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.38, 0.121, 0.566, 0.995] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1387, "options": [ "A. talk to (e.g., self, a person, a group) [0.461, 0.210, 0.544, 0.739] -> cut [0.430, 0.126, 0.613, 0.927] -> walk [0.428, 0.054, 0.550, 1.000]", "B. crouch/kneel [0.599, 0.017, 0.821, 0.918] -> talk to (e.g., self, a person, a group) [0.615, 0.017, 0.693, 0.729] -> walk [0.653, 0.055, 0.630, 0.744]", "C. walk [0.46, 0.143, 0.72, 0.92] -> carry/hold (an object) [0.46, 0.143, 0.72, 0.92] -> talk to (e.g., self, a person, a group) [0.46, 0.143, 0.72, 0.92]", "D. talk to (e.g., self, a person, a group) [0.564, 0.112, 0.773, 0.851] -> give/serve (an object) to (a person) [0.553, 0.039, 0.551, 0.848] -> walk [0.400, 0.318, 0.745, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.46, 0.143, 0.72, 0.92] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1399, "options": [ "A. stand, listen to (a person), watch (a person)", "B. walk, listen to (a person)", "C. walk, talk to (e.g., self, a person, a group)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.362, 0.193, 0.682, 0.982] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1428, "options": [ "A. [0.068, 0.275, 0.776, 1.0]", "B. [0.01, 0.048, 0.54, 0.7]", "C. [0.185, 0.014, 0.907, 0.984]", "D. [0.0, 0.0, 0.752, 0.978]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1534, "options": [ "A. [0.348, 0.0, 0.906, 0.688]", "B. [0.127, 0.0, 1.0, 0.982]", "C. [0.59, 0.149, 0.881, 0.993]", "D. [0.0, 0.089, 0.805, 1.0]" ], "question": "Where is the person currently performing the smoke located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1569, "options": [ "A. drive (e.g., a car, a truck)", "B. carry/hold (an object)", "C. play board game", "D. read" ], "question": "What action is the person currently in the [0.452, 0.077, 0.897, 0.991] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1574, "options": [ "A. [0.422, 0.372, 0.881, 1.0]", "B. [0.182, 0.379, 0.437, 0.997]", "C. [0.141, 0.078, 0.634, 1.0]", "D. [0.036, 0.0, 0.429, 0.99]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1679, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.441, 0.07, 0.656, 0.712] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1689, "options": [ "A. [0.571, 0.366, 0.9, 0.41]", "B. [0.434, 0.067, 0.652, 0.653]", "C. [0.149, 0.0, 0.504, 0.693]", "D. [0.54, 0.083, 0.828, 0.976]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1770, "options": [ "A. sit, listen to (a person), watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. walk, talk to (e.g., self, a person, a group)", "D. listen to (a person), lie/sleep, watch (a person)" ], "question": "What action is the person at the location [0.0, 0.023, 0.656, 0.99] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 902, 1770 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 910, "options": [ "A. [0.743, 0.107, 0.91, 0.546]", "B. [0.687, 0.38, 0.946, 1.0]", "C. [0.645, 0.213, 0.874, 0.801]", "D. [0.137, 0.047, 0.701, 0.992]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 928, "options": [ "A. [0.0, 0.209, 0.374, 0.786]", "B. [0.13, 0.181, 0.554, 0.801]", "C. [0.0, 0.135, 0.339, 0.543]", "D. [0.196, 0.055, 0.814, 0.996]" ], "question": "Where was the person currently performing the drive (e.g., a car, a truck) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1031, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1088, "options": [ "A. lie/sleep -> hand wave", "B. lie/sleep -> answer phone", "C. carry/hold (an object) -> lie/sleep", "D. answer phone -> push (another person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1131, "options": [ "A. walk, carry/hold (an object), give/serve (an object) to (a person), watch (a person)", "B. sit, drive (e.g., a car, a truck), listen to (a person)", "C. sit, drive (e.g., a car, a truck)", "D. sit, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.122, 0.063, 0.316, 0.975] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1194, "options": [ "A. touch (an object) [0.134, 0.051, 0.424, 0.835] -> answer phone [0.164, 0.11, 0.403, 0.771] -> carry/hold (an object) [0.164, 0.11, 0.403, 0.771]", "B. touch (an object) [0.170, 0.146, 0.342, 0.951] -> answer phone [0.000, 0.310, 0.455, 0.594] -> press [0.032, 0.277, 0.492, 0.810]", "C. answer phone [0.262, 0.000, 0.524, 0.816] -> carry/hold (an object) [0.355, 0.049, 0.268, 0.843] -> lift (a person) [0.150, 0.144, 0.363, 0.877]", "D. answer phone [0.182, 0.249, 0.383, 0.734] -> carry/hold (an object) [0.170, 0.141, 0.497, 0.669] -> pull (an object) [0.203, 0.000, 0.300, 0.768]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.164, 0.11, 0.403, 0.771] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1195, "options": [ "A. answer phone and touch (an object)", "B. fight/hit (a person)", "C. touch (an object)", "D. crawl" ], "question": "What action is the person currently in the [0.536, 0.667, 0.604, 0.973] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1200, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.129, 0.084, 0.416, 0.753] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1207, "options": [ "A. answer phone -> push (an object)", "B. touch (an object) -> dance", "C. kiss (a person) -> answer phone", "D. answer phone -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1217, "options": [ "A. 6", "B. 5", "C. 3", "D. 4" ], "question": "How many people were performing the watch (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1255, "options": [ "A. paint [0.725, 0.030, 0.699, 0.786] -> bend/bow (at the waist) [0.569, 0.040, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.524, 0.215, 0.930, 0.787]", "B. touch (an object) [0.791, 0.090, 1.000, 0.818] -> press [0.762, 0.057, 0.823, 1.000] -> talk to (e.g., self, a person, a group) [0.697, 0.247, 0.844, 0.814]", "C. touch (an object) [0.680, 0.000, 0.862, 1.000] -> play with kids [0.613, 0.140, 0.706, 1.000] -> bend/bow (at the waist) [0.460, 0.252, 0.778, 0.848]", "D. bend/bow (at the waist) [0.636, 0.137, 0.876, 0.975] -> touch (an object) [0.636, 0.137, 0.876, 0.975] -> talk to (e.g., self, a person, a group) [0.636, 0.137, 0.876, 0.975]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.636, 0.137, 0.876, 0.975] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1255, "options": [ "A. talk to (e.g., self, a person, a group) -> answer phone -> touch (an object)", "B. talk to (e.g., self, a person, a group) -> bend/bow (at the waist) -> touch (an object)", "C. bend/bow (at the waist) -> touch (an object) -> extract", "D. bend/bow (at the waist) -> touch (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1278, "options": [ "A. drive (e.g., a car, a truck) [0.212, 0.003, 0.799, 0.928] -> talk to (e.g., self, a person, a group) [0.212, 0.003, 0.799, 0.928] -> carry/hold (an object) [0.243, 0.0, 0.833, 0.966]", "B. carry/hold (an object) [0.193, 0.136, 0.952, 1.000] -> talk to (e.g., self, a person, a group) [0.353, 0.060, 0.634, 0.937] -> chop [0.022, 0.199, 0.642, 1.000]", "C. carry/hold (an object) [0.258, 0.165, 0.764, 0.788] -> row boat [0.050, 0.154, 0.774, 0.863] -> drive (e.g., a car, a truck) [0.127, 0.000, 0.628, 0.973]", "D. talk to (e.g., self, a person, a group) [0.356, 0.174, 0.929, 0.954] -> drive (e.g., a car, a truck) [0.086, 0.131, 0.841, 0.895] -> paint [0.351, 0.055, 0.846, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.243, 0.0, 0.833, 0.966] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1293, "options": [ "A. fishing -> touch (an object)", "B. chop -> answer phone", "C. answer phone -> touch (an object)", "D. touch (an object) -> hug (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1357, "options": [ "A. dig [0.433, 0.137, 0.837, 1.000] -> bend/bow (at the waist) [0.451, 0.244, 0.956, 0.792] -> talk to (e.g., self, a person, a group) [0.515, 0.158, 0.830, 1.000]", "B. talk to (e.g., self, a person, a group) [0.623, 0.074, 0.843, 0.868] -> fall down [0.502, 0.162, 0.847, 0.985] -> touch (an object) [0.353, 0.026, 0.864, 0.932]", "C. talk to (e.g., self, a person, a group) [0.451, 0.248, 0.803, 0.977] -> bend/bow (at the waist) [0.383, 0.114, 0.819, 0.98] -> touch (an object) [0.383, 0.114, 0.819, 0.98]", "D. kiss (a person) [0.528, 0.083, 0.838, 1.000] -> bend/bow (at the waist) [0.304, 0.286, 0.645, 0.976] -> touch (an object) [0.233, 0.123, 0.670, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.383, 0.114, 0.819, 0.98] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1380, "options": [ "A. [0.591, 0.132, 0.763, 0.584]", "B. [0.573, 0.652, 0.692, 0.991]", "C. [0.335, 0.34, 0.855, 0.985]", "D. [0.308, 0.213, 0.876, 0.802]" ], "question": "Where was the person currently performing the drive (e.g., a car, a truck) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1401, "options": [ "A. [0.122, 0.094, 0.347, 0.734]", "B. [0.148, 0.04, 0.759, 0.989]", "C. [0.153, 0.32, 0.827, 1.0]", "D. [0.446, 0.083, 0.938, 0.832]" ], "question": "Where is the person currently performing the ride (e.g., a bike, a car, a horse) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1408, "options": [ "A. ride (e.g., a bike, a car, a horse) -> hand clap", "B. ride (e.g., a bike, a car, a horse) -> answer phone", "C. answer phone -> play board game", "D. touch (an object) -> ride (e.g., a bike, a car, a horse)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1412, "options": [ "A. [0.3, 0.012, 0.88, 0.719]", "B. [0.0, 0.121, 0.689, 1.0]", "C. [0.505, 0.0, 0.875, 1.0]", "D. [0.273, 0.028, 0.724, 1.0]" ], "question": "Where is the person currently performing the answer phone located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1452, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1466, "options": [ "A. [0.153, 0.061, 0.709, 1.0]", "B. [0.63, 0.261, 0.878, 0.812]", "C. [0.453, 0.389, 0.652, 0.975]", "D. [0.66, 0.0, 0.712, 0.63]" ], "question": "Where was the person currently performing the ride (e.g., a bike, a car, a horse) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1626, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.158, 0.022, 0.58, 0.955] location in the current frame performing the drive (e.g., a car, a truck)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1648, "options": [ "A. carry/hold (an object) [0.608, 0.184, 0.857, 0.629] -> ride (e.g., a bike, a car, a horse) [0.608, 0.184, 0.857, 0.629] -> talk to (e.g., self, a person, a group) [0.608, 0.184, 0.857, 0.629]", "B. push (another person) [0.512, 0.119, 0.827, 0.480] -> carry/hold (an object) [0.575, 0.352, 0.769, 0.651] -> talk to (e.g., self, a person, a group) [0.760, 0.157, 0.985, 0.581]", "C. talk to (e.g., self, a person, a group) [0.707, 0.033, 0.777, 0.468] -> carry/hold (an object) [0.684, 0.145, 0.689, 0.687] -> jump/leap [0.474, 0.166, 0.960, 0.566]", "D. ride (e.g., a bike, a car, a horse) [0.583, 0.203, 0.812, 0.796] -> hit (an object) [0.565, 0.024, 0.726, 0.445] -> carry/hold (an object) [0.744, 0.059, 0.938, 0.693]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.608, 0.184, 0.857, 0.629] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1712, "options": [ "A. sit, drive (e.g., a car, a truck), listen to (a person)", "B. sit, drive (e.g., a car, a truck), talk to (e.g., self, a person, a group)", "C. stand, listen to (a person), watch (a person)", "D. walk, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.165, 0.071, 0.222, 0.355] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1787, "options": [ "A. 7", "B. 5", "C. 3", "D. 4" ], "question": "How many people were performing the stand in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1787 ]

[ { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 943, "options": [ "A. talk to (e.g., self, a person, a group) [0.662, 0.246, 0.888, 0.969] -> walk [0.715, 0.221, 0.962, 0.976] -> carry/hold (an object) [0.715, 0.221, 0.962, 0.976]", "B. carry/hold (an object) [0.516, 0.267, 0.928, 1.000] -> stir [0.754, 0.350, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.514, 0.436, 0.789, 0.995]", "C. talk to (e.g., self, a person, a group) [0.676, 0.401, 0.844, 0.784] -> turn (e.g., a screwdriver) [0.876, 0.170, 0.806, 0.874] -> carry/hold (an object) [0.690, 0.405, 1.000, 0.822]", "D. carry/hold (an object) [0.544, 0.168, 0.927, 1.000] -> talk to (e.g., self, a person, a group) [0.585, 0.133, 0.731, 1.000] -> stir [0.641, 0.242, 1.000, 0.922]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.715, 0.221, 0.962, 0.976] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1005, "options": [ "A. sit, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, touch (an object), listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "C. sit, text on/look at a cellphone, touch (an object), listen to (a person)", "D. sit, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.112, 0.341, 0.483, 0.798] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1045, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.014, 0.014, 0.931, 0.992] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1099, "options": [ "A. [0.216, 0.477, 0.596, 1.0]", "B. [0.383, 0.277, 0.29, 0.54]", "C. [0.508, 0.297, 0.772, 0.861]", "D. [0.196, 0.351, 0.459, 0.807]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1104, "options": [ "A. [0.009, 0.03, 0.977, 0.993]", "B. [0.055, 0.0, 1.0, 0.696]", "C. [0.296, 0.0, 1.0, 1.0]", "D. [0.218, 0.088, 0.405, 0.79]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1132, "options": [ "A. crouch/kneel [0.482, 0.237, 1.000, 0.959] -> carry/hold (an object) [0.587, 0.249, 1.000, 0.846] -> bend/bow (at the waist) [0.397, 0.089, 0.935, 0.914]", "B. carry/hold (an object) [0.286, 0.110, 0.688, 0.808] -> carry/hold (an object) [0.391, 0.149, 1.000, 1.000] -> bend/bow (at the waist) [0.645, 0.326, 0.954, 1.000]", "C. bend/bow (at the waist) [0.578, 0.320, 1.000, 0.874] -> give/serve (an object) to (a person) [0.522, 0.392, 0.920, 0.800] -> dig [0.636, 0.021, 0.702, 0.882]", "D. bend/bow (at the waist) [0.449, 0.208, 0.882, 0.987] -> carry/hold (an object) [0.449, 0.208, 0.882, 0.987] -> give/serve (an object) to (a person) [0.449, 0.208, 0.882, 0.987]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.449, 0.208, 0.882, 0.987] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1166, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1222, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "B. sit, watch (a person)", "C. sit, talk to (e.g., self, a person, a group), watch (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.168, 0.0, 0.385, 0.994] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1290, "options": [ "A. [0.57, 0.0, 1.0, 0.706]", "B. [0.168, 0.221, 0.521, 0.743]", "C. [0.791, 0.349, 0.99, 1.0]", "D. [0.584, 0.067, 0.992, 0.975]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1307, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.469, 0.0, 0.815, 0.894] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1307, "options": [ "A. [0.486, 0.262, 1.0, 1.0]", "B. [0.691, 0.0, 1.0, 0.732]", "C. [0.469, 0.0, 0.815, 0.894]", "D. [0.3, 0.358, 0.698, 0.912]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1313, "options": [ "A. carry/hold (an object) [0.310, 0.031, 0.625, 1.000] -> exit [0.362, 0.321, 0.389, 0.997] -> walk [0.424, 0.055, 0.436, 1.000]", "B. walk [0.401, 0.205, 0.413, 0.817] -> carry/hold (an object) [0.207, 0.043, 0.533, 1.000] -> push (another person) [0.407, 0.321, 0.513, 0.780]", "C. talk to (e.g., self, a person, a group) [0.345, 0.168, 0.484, 0.828] -> walk [0.285, 0.07, 0.538, 1.0] -> carry/hold (an object) [0.285, 0.07, 0.538, 1.0]", "D. talk to (e.g., self, a person, a group) [0.355, 0.318, 0.339, 0.747] -> walk [0.291, 0.000, 0.514, 0.806] -> pull (an object) [0.461, 0.099, 0.417, 0.925]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.285, 0.07, 0.538, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1404, "options": [ "A. [0.275, 0.292, 0.597, 0.874]", "B. [0.342, 0.584, 0.424, 0.909]", "C. [0.495, 0.065, 0.989, 0.93]", "D. [0.378, 0.005, 0.589, 0.596]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1448, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1563, "options": [ "A. [0.33, 0.082, 1.0, 1.0]", "B. [0.515, 0.086, 0.695, 0.642]", "C. [0.046, 0.005, 0.933, 0.986]", "D. [0.016, 0.0, 0.644, 0.766]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1663, "options": [ "A. [0.025, 0.145, 0.583, 0.977]", "B. [0.152, 0.307, 0.79, 1.0]", "C. [0.305, 0.2, 0.663, 0.713]", "D. [0.538, 0.087, 0.705, 0.623]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1678, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1748, "options": [ "A. touch (an object) [0.662, 0.341, 1.000, 0.744] -> play with kids [0.605, 0.340, 1.000, 0.770] -> carry/hold (an object) [0.374, 0.301, 1.000, 0.930]", "B. carry/hold (an object) [0.615, 0.338, 0.765, 0.955] -> kick (an object) [0.383, 0.395, 0.813, 0.790] -> touch (an object) [0.456, 0.403, 0.880, 0.710]", "C. talk to (e.g., self, a person, a group) [0.549, 0.28, 0.898, 0.81] -> carry/hold (an object) [0.548, 0.276, 0.901, 0.799] -> touch (an object) [0.548, 0.276, 0.901, 0.799]", "D. talk to (e.g., self, a person, a group) [0.441, 0.179, 0.752, 0.907] -> extract [0.551, 0.198, 0.864, 0.788] -> touch (an object) [0.615, 0.273, 1.000, 0.614]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.548, 0.276, 0.901, 0.799] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1751, "options": [ "A. talk to (e.g., self, a person, a group) [0.222, 0.260, 0.655, 0.854] -> hug (a person) [0.328, 0.359, 0.611, 0.905] -> carry/hold (an object) [0.287, 0.209, 0.713, 0.782]", "B. talk to (e.g., self, a person, a group) [0.319, 0.440, 0.378, 0.886] -> stir [0.168, 0.229, 0.672, 0.805] -> carry/hold (an object) [0.091, 0.252, 0.366, 0.853]", "C. talk to (e.g., self, a person, a group) [0.063, 0.330, 0.440, 0.595] -> touch (an object) [0.185, 0.151, 0.556, 0.779] -> crouch/kneel [0.141, 0.388, 0.565, 0.650]", "D. carry/hold (an object) [0.172, 0.321, 0.551, 0.737] -> touch (an object) [0.172, 0.321, 0.551, 0.737] -> talk to (e.g., self, a person, a group) [0.172, 0.321, 0.551, 0.737]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.172, 0.321, 0.551, 0.737] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1763, "options": [ "A. carry/hold (an object) and talk to (e.g., self, a person, a group)", "B. cut", "C. lift (a person)", "D. text on/look at a cellphone" ], "question": "What action is the person currently in the [0.176, 0.198, 0.516, 0.757] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1773, "options": [ "A. [0.845, 0.443, 1.0, 0.543]", "B. [0.582, 0.268, 0.957, 0.833]", "C. [0.292, 0.369, 1.0, 0.972]", "D. [0.028, 0.064, 1.0, 0.992]" ], "question": "Where is the person currently performing the touch (an object) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1773 ]

[ { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 914, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 944, "options": [ "A. [0.0, 0.084, 0.378, 1.0]", "B. [0.325, 0.078, 0.977, 1.0]", "C. [0.032, 0.008, 0.674, 0.927]", "D. [0.079, 0.217, 0.814, 0.985]" ], "question": "Where was the person currently performing the sail boat in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 967, "options": [ "A. sit, listen to (a person)", "B. stand, watch (a person)", "C. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. sit, ride (e.g., a bike, a car, a horse)" ], "question": "What action is the person at the location [0.007, 0.047, 0.366, 0.991] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1080, "options": [ "A. [0.016, 0.0, 1.0, 0.702]", "B. [0.337, 0.0, 1.0, 0.716]", "C. [0.089, 0.253, 0.4, 0.969]", "D. [0.308, 0.084, 0.989, 0.984]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1149, "options": [ "A. [0.103, 0.051, 0.67, 1.0]", "B. [0.307, 0.002, 0.542, 0.514]", "C. [0.354, 0.172, 0.731, 0.763]", "D. [0.205, 0.44, 0.802, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1164, "options": [ "A. watch (e.g., TV)", "B. exit", "C. ride (e.g., a bike, a car, a horse)", "D. drive (e.g., a car, a truck)" ], "question": "What action is the person currently in the [0.456, 0.013, 0.987, 0.98] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1177, "options": [ "A. [0.217, 0.003, 0.941, 0.959]", "B. [0.119, 0.274, 1.0, 1.0]", "C. [0.006, 0.0, 0.827, 1.0]", "D. [0.625, 0.012, 1.0, 0.99]" ], "question": "Where was the person currently performing the ride (e.g., a bike, a car, a horse) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1187, "options": [ "A. [0.343, 0.25, 0.801, 1.0]", "B. [0.046, 0.009, 0.741, 0.916]", "C. [0.518, 0.001, 0.998, 1.0]", "D. [0.563, 0.0, 0.76, 0.737]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1219, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.0, 0.347, 0.995] location in the current frame performing the drive (e.g., a car, a truck)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1251, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1259, "options": [ "A. [0.418, 0.109, 0.989, 1.0]", "B. [0.0, 0.0, 0.595, 0.687]", "C. [0.259, 0.0, 0.88, 0.716]", "D. [0.047, 0.0, 0.787, 0.982]" ], "question": "Where is the person currently performing the ride (e.g., a bike, a car, a horse) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1450, "options": [ "A. [0.355, 0.076, 0.982, 0.979]", "B. [0.408, 0.311, 0.494, 1.0]", "C. [0.245, 0.167, 0.058, 0.794]", "D. [0.138, 0.455, 0.286, 0.992]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1487, "options": [ "A. [0.456, 0.338, 0.712, 0.991]", "B. [0.0, 0.0, 0.661, 0.752]", "C. [0.353, 0.0, 0.703, 1.0]", "D. [0.114, 0.169, 0.68, 0.994]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1503, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1600, "options": [ "A. [0.446, 0.0, 0.842, 0.914]", "B. [0.136, 0.472, 0.393, 0.988]", "C. [0.075, 0.0, 0.573, 0.738]", "D. [0.16, 0.0, 0.776, 0.995]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1676, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1676 ]

[ { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1061, "options": [ "A. [0.0, 0.047, 0.581, 0.985]", "B. [0.111, 0.038, 0.825, 0.979]", "C. [0.0, 0.317, 1.0, 1.0]", "D. [0.389, 0.0, 0.891, 0.788]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1095, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1151, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.505, 0.0, 1.0, 0.987] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1191, "options": [ "A. [0.017, 0.366, 0.279, 0.992]", "B. [0.205, 0.0, 0.871, 0.635]", "C. [0.0, 0.01, 0.917, 0.935]", "D. [0.0, 0.166, 0.673, 1.0]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1217, "options": [ "A. [0.071, 0.263, 0.536, 1.0]", "B. [0.319, 0.145, 0.826, 0.703]", "C. [0.309, 0.306, 0.767, 0.983]", "D. [0.273, 0.16, 0.53, 0.859]" ], "question": "Where was the person currently performing the bend/bow (at the waist) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1316, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.237, 0.508, 0.287, 0.835] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1326, "options": [ "A. [0.834, 0.0, 0.509, 0.461]", "B. [0.139, 0.216, 0.438, 0.995]", "C. [0.618, 0.189, 0.733, 0.626]", "D. [0.761, 0.293, 0.796, 0.831]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1438, "options": [ "A. [0.07, 0.156, 0.504, 0.974]", "B. [0.298, 0.19, 0.945, 0.748]", "C. [0.541, 0.29, 0.887, 0.921]", "D. [0.768, 0.498, 1.0, 0.765]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1453, "options": [ "A. walk and talk to (e.g., self, a person, a group)", "B. shovel", "C. play with pets", "D. carry/hold (an object)" ], "question": "What action is the person currently in the [0.01, 0.02, 0.644, 0.98] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1481, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.141, 0.197, 0.487, 0.986] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1515, "options": [ "A. [0.098, 0.161, 0.404, 1.0]", "B. [0.249, 0.083, 0.4, 0.606]", "C. [0.331, 0.217, 0.698, 1.0]", "D. [0.0, 0.085, 0.139, 1.0]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1525, "options": [ "A. [0.563, 0.251, 0.853, 0.977]", "B. [0.0, 0.0, 0.564, 0.727]", "C. [0.0, 0.0, 0.16, 1.0]", "D. [0.082, 0.186, 0.386, 0.994]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1538, "options": [ "A. [0.224, 0.0, 0.675, 0.785]", "B. [0.306, 0.277, 0.541, 1.0]", "C. [0.132, 0.231, 0.387, 0.98]", "D. [0.469, 0.053, 0.762, 0.989]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1551, "options": [ "A. [0.475, 0.059, 0.754, 0.978]", "B. [0.524, 0.322, 0.877, 0.992]", "C. [0.0, 0.455, 0.333, 0.988]", "D. [0.329, 0.0, 0.615, 0.897]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1768, "options": [ "A. sit, listen to (a person)", "B. stand, listen to (a person), watch (a person)", "C. sit, carry/hold (an object), listen to (a person)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.142, 0.27, 0.456, 0.721] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 941, 1768 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 903, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 903, "options": [ "A. talk to (e.g., self, a person, a group) [0.754, 0.550, 0.824, 0.889] -> carry/hold (an object) [0.637, 0.603, 0.872, 0.902] -> hand shake [0.874, 0.451, 0.621, 0.773]", "B. talk to (e.g., self, a person, a group) [0.573, 0.507, 0.880, 0.565] -> kiss (a person) [0.562, 0.433, 0.816, 0.902] -> crouch/kneel [0.865, 0.619, 0.822, 0.770]", "C. carry/hold (an object) [0.674, 0.531, 0.741, 0.751] -> crouch/kneel [0.674, 0.531, 0.741, 0.751] -> talk to (e.g., self, a person, a group) [0.674, 0.531, 0.741, 0.751]", "D. carry/hold (an object) [0.754, 0.630, 0.580, 0.587] -> talk to (e.g., self, a person, a group) [0.508, 0.573, 0.909, 0.707] -> crouch/kneel [0.659, 0.332, 0.733, 0.916]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.674, 0.531, 0.741, 0.751] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 939, "options": [ "A. ride (e.g., a bike, a car, a horse) -> kick (an object)", "B. carry/hold (an object) -> row boat", "C. ride (e.g., a bike, a car, a horse) -> carry/hold (an object)", "D. eat -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 943, "options": [ "A. [0.816, 0.361, 0.85, 0.501]", "B. [0.077, 0.021, 0.779, 0.989]", "C. [1.0, 0.207, 0.925, 0.5]", "D. [0.539, 0.092, 1.0, 0.354]" ], "question": "Where is the person currently performing the run/jog located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1008, "options": [ "A. [0.0, 0.158, 0.59, 0.972]", "B. [0.663, 0.161, 0.996, 0.943]", "C. [0.614, 0.259, 0.745, 1.0]", "D. [0.588, 0.0, 0.716, 0.829]" ], "question": "Where was the person currently performing the read in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1020, "options": [ "A. talk to (e.g., self, a person, a group) [0.702, 0.178, 0.987, 0.941] -> carry/hold (an object) [0.706, 0.168, 0.989, 0.938] -> read [0.706, 0.168, 0.989, 0.938]", "B. shoot [0.727, 0.345, 1.000, 0.960] -> talk to (e.g., self, a person, a group) [0.614, 0.113, 0.957, 0.745] -> read [0.758, 0.143, 0.995, 1.000]", "C. carry/hold (an object) [0.740, 0.083, 0.921, 1.000] -> talk to (e.g., self, a person, a group) [0.899, 0.196, 0.989, 0.840] -> ride (e.g., a bike, a car, a horse) [0.670, 0.082, 1.000, 0.973]", "D. fight/hit (a person) [0.592, 0.102, 1.000, 0.964] -> carry/hold (an object) [0.643, 0.015, 0.861, 0.867] -> talk to (e.g., self, a person, a group) [0.799, 0.121, 0.987, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.706, 0.168, 0.989, 0.938] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1121, "options": [ "A. drive (e.g., a car, a truck) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> brush teeth", "C. talk to (e.g., self, a person, a group) -> give/serve (an object) to (a person)", "D. clink glass -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1123, "options": [ "A. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group)", "B. row boat -> talk to (e.g., self, a person, a group)", "C. ride (e.g., a bike, a car, a horse) -> dance", "D. take (an object) from (a person) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1134, "options": [ "A. talk to (e.g., self, a person, a group) -> drive (e.g., a car, a truck)", "B. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group)", "C. shoot -> ride (e.g., a bike, a car, a horse)", "D. talk to (e.g., self, a person, a group) -> catch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1145, "options": [ "A. hit (an object) [0.000, 0.102, 0.605, 1.000] -> carry/hold (an object) [0.013, 0.085, 0.444, 0.971] -> bend/bow (at the waist) [0.282, 0.254, 0.279, 0.856]", "B. talk to (e.g., self, a person, a group) [0.231, 0.049, 0.444, 1.000] -> swim [0.242, 0.260, 0.436, 0.823] -> bend/bow (at the waist) [0.089, 0.159, 0.372, 0.862]", "C. bend/bow (at the waist) [0.086, 0.193, 0.406, 0.97] -> carry/hold (an object) [0.086, 0.193, 0.406, 0.97] -> talk to (e.g., self, a person, a group) [0.086, 0.193, 0.406, 0.97]", "D. talk to (e.g., self, a person, a group) [0.000, 0.220, 0.312, 1.000] -> drive (e.g., a car, a truck) [0.269, 0.007, 0.501, 1.000] -> carry/hold (an object) [0.146, 0.010, 0.590, 0.791]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.086, 0.193, 0.406, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1156, "options": [ "A. bend/bow (at the waist) [0.181, 0.167, 0.411, 0.862] -> carry/hold (an object) [0.012, 0.344, 0.055, 0.838] -> talk to (e.g., self, a person, a group) [0.000, 0.154, 0.187, 0.863]", "B. bend/bow (at the waist) [0.042, 0.166, 0.225, 0.733] -> carry/hold (an object) [0.042, 0.166, 0.225, 0.733] -> talk to (e.g., self, a person, a group) [0.042, 0.166, 0.225, 0.733]", "C. talk to (e.g., self, a person, a group) [0.000, 0.150, 0.176, 0.634] -> carry/hold (an object) [0.084, 0.346, 0.197, 0.746] -> touch (an object) [0.191, 0.215, 0.255, 0.775]", "D. brush teeth [0.000, 0.297, 0.312, 0.850] -> talk to (e.g., self, a person, a group) [0.020, 0.045, 0.127, 0.851] -> bend/bow (at the waist) [0.147, 0.000, 0.045, 0.660]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.042, 0.166, 0.225, 0.733] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1298, "options": [ "A. sit, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. walk, carry/hold (an object), give/serve (an object) to (a person), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.0, 0.042, 0.452, 0.969] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1368, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1379, "options": [ "A. [0.0, 0.293, 0.301, 1.0]", "B. [0.052, 0.0, 0.767, 1.0]", "C. [0.698, 0.479, 0.934, 0.973]", "D. [0.039, 0.065, 0.48, 0.957]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1380, "options": [ "A. sail boat -> carry/hold (an object)", "B. take (an object) from (a person) -> carry/hold (an object)", "C. hand shake -> carry/hold (an object)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1490, "options": [ "A. 2", "B. 5", "C. 6", "D. 3" ], "question": "How many people in the current frame are performing the action: walk?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1513, "options": [ "A. [0.432, 0.14, 0.906, 0.959]", "B. [0.331, 0.237, 0.457, 0.938]", "C. [0.358, 0.043, 0.494, 0.784]", "D. [0.454, 0.363, 0.648, 0.986]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1564, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.155, 0.125, 0.974, 0.946] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1567, "options": [ "A. [0.0, 0.336, 0.696, 1.0]", "B. [0.305, 0.17, 0.872, 0.973]", "C. [0.016, 0.094, 0.556, 0.96]", "D. [0.212, 0.0, 0.356, 0.718]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1692, "options": [ "A. 5", "B. 2", "C. 3", "D. 4" ], "question": "How many people in the current frame are performing the action: walk?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1708, "options": [ "A. [0.377, 0.131, 0.53, 1.0]", "B. [0.542, 0.217, 0.736, 0.953]", "C. [0.205, 0.003, 0.454, 0.952]", "D. [0.734, 0.415, 0.935, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1730, "options": [ "A. 3", "B. 6", "C. 4", "D. 5" ], "question": "How many people were performing the walk in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1757, "options": [ "A. talk to (e.g., self, a person, a group) [0.417, 0.241, 0.665, 0.95] -> walk [0.475, 0.227, 0.733, 0.95] -> grab (a person) [0.475, 0.227, 0.733, 0.95]", "B. walk [0.380, 0.314, 0.922, 1.000] -> talk to (e.g., self, a person, a group) [0.225, 0.335, 0.705, 0.862] -> grab (a person) [0.633, 0.417, 0.813, 0.906]", "C. catch (an object) [0.402, 0.349, 0.688, 0.800] -> grab (a person) [0.482, 0.181, 0.731, 0.798] -> talk to (e.g., self, a person, a group) [0.532, 0.256, 0.552, 0.960]", "D. grab (a person) [0.373, 0.078, 0.648, 0.842] -> paint [0.431, 0.347, 0.759, 0.937] -> walk [0.484, 0.076, 0.603, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.475, 0.227, 0.733, 0.95] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1766, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.132, 0.084, 0.598, 0.967] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1778, "options": [ "A. carry/hold (an object)", "B. walk and talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group)", "D. put down" ], "question": "What action is the person currently in the [0.442, 0.078, 0.951, 0.948] location likely to do next?", "sub_answer_type": "Action Anticipation" } ]

[ 900, 1778 ]

[ { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 916, "options": [ "A. answer phone -> smoke", "B. answer phone -> cook", "C. kick (a person) -> answer phone", "D. answer phone -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 934, "options": [ "A. touch (an object) [0.35, 0.175, 0.538, 0.938] -> walk [0.342, 0.152, 0.517, 0.93] -> enter [0.342, 0.152, 0.517, 0.93]", "B. walk [0.359, 0.319, 0.421, 0.884] -> touch (an object) [0.390, 0.094, 0.691, 0.740] -> cook [0.395, 0.241, 0.649, 0.881]", "C. enter [0.380, 0.178, 0.627, 0.988] -> touch (an object) [0.543, 0.340, 0.603, 0.840] -> climb (e.g., a mountain) [0.416, 0.339, 0.405, 0.764]", "D. touch (an object) [0.216, 0.000, 0.606, 0.998] -> enter [0.406, 0.103, 0.477, 0.736] -> cut [0.168, 0.280, 0.559, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.342, 0.152, 0.517, 0.93] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 947, "options": [ "A. hand clap", "B. paint", "C. walk and answer phone", "D. walk" ], "question": "What action is the person currently in the [0.334, 0.123, 0.91, 0.972] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 956, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.241, 0.059, 0.925, 0.95] location in the current frame performing the answer phone?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 985, "options": [ "A. [0.321, 0.084, 0.918, 0.932]", "B. [0.013, 0.071, 0.906, 0.954]", "C. [0.0, 0.133, 0.64, 0.778]", "D. [0.0, 0.365, 0.768, 0.954]" ], "question": "Where was the person currently performing the bend/bow (at the waist) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1004, "options": [ "A. [0.083, 0.102, 0.842, 0.965]", "B. [0.0, 0.39, 0.563, 0.997]", "C. [0.054, 0.062, 0.833, 0.681]", "D. [0.397, 0.107, 0.987, 0.946]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1102, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.108, 0.062, 0.957, 0.959] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1111, "options": [ "A. [0.652, 0.318, 0.746, 0.746]", "B. [0.057, 0.0, 0.664, 0.864]", "C. [0.138, 0.039, 0.945, 0.948]", "D. [0.213, 0.324, 1.0, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1143, "options": [ "A. talk to (e.g., self, a person, a group) -> take (an object) from (a person)", "B. lift/pick up -> work on a computer", "C. talk to (e.g., self, a person, a group) -> drink", "D. talk to (e.g., self, a person, a group) -> work on a computer" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1165, "options": [ "A. answer phone [0.000, 0.180, 0.471, 0.805] -> swim [0.373, 0.108, 0.413, 0.855] -> carry/hold (an object) [0.011, 0.129, 0.738, 0.810]", "B. carry/hold (an object) [0.257, 0.118, 0.474, 0.623] -> touch (an object) [0.331, 0.228, 0.544, 0.630] -> swim [0.285, 0.381, 0.653, 0.715]", "C. touch (an object) [0.202, 0.248, 0.502, 0.704] -> answer phone [0.176, 0.262, 0.554, 0.721] -> carry/hold (an object) [0.176, 0.262, 0.554, 0.721]", "D. put down [0.161, 0.155, 0.379, 0.541] -> carry/hold (an object) [0.243, 0.381, 0.439, 0.657] -> answer phone [0.133, 0.273, 0.720, 0.878]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.176, 0.262, 0.554, 0.721] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1171, "options": [ "A. [0.422, 0.271, 0.877, 0.905]", "B. [0.453, 0.116, 0.915, 0.452]", "C. [0.27, 0.248, 0.73, 0.7]", "D. [0.0, 0.249, 0.848, 0.967]" ], "question": "Where was the person currently performing the write in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1175, "options": [ "A. touch (an object) [0.126, 0.234, 0.449, 0.618] -> write [0.126, 0.234, 0.449, 0.618] -> talk to (e.g., self, a person, a group) [0.126, 0.234, 0.449, 0.618]", "B. lift/pick up [0.072, 0.327, 0.451, 0.604] -> talk to (e.g., self, a person, a group) [0.000, 0.413, 0.350, 0.817] -> touch (an object) [0.010, 0.254, 0.592, 0.678]", "C. touch (an object) [0.319, 0.187, 0.389, 0.797] -> play board game [0.237, 0.289, 0.522, 0.436] -> talk to (e.g., self, a person, a group) [0.311, 0.295, 0.509, 0.687]", "D. touch (an object) [0.061, 0.394, 0.603, 0.588] -> talk to (e.g., self, a person, a group) [0.000, 0.408, 0.590, 0.596] -> catch (an object) [0.054, 0.264, 0.351, 0.447]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.126, 0.234, 0.449, 0.618] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1185, "options": [ "A. [0.436, 0.0, 0.338, 0.343]", "B. [0.008, 0.098, 0.605, 0.943]", "C. [0.121, 0.41, 0.148, 0.282]", "D. [0.187, 0.214, 0.435, 0.485]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1197, "options": [ "A. touch (an object) -> answer phone -> carry/hold (an object) -> write -> talk to (e.g., self, a person, a group)", "B. dig -> write -> answer phone -> touch (an object) -> talk to (e.g., self, a person, a group)", "C. touch (an object) -> answer phone -> carry/hold (an object) -> play board game -> talk to (e.g., self, a person, a group)", "D. answer phone -> hand clap -> talk to (e.g., self, a person, a group) -> write -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1227, "options": [ "A. [0.301, 0.0, 0.739, 0.756]", "B. [0.153, 0.152, 0.792, 0.965]", "C. [0.355, 0.202, 0.907, 1.0]", "D. [0.006, 0.08, 0.478, 0.964]" ], "question": "Where was the person currently performing the work on a computer in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1298, "options": [ "A. eat [0.331, 0.049, 0.623, 1.000] -> walk [0.370, 0.122, 0.386, 0.914] -> talk to (e.g., self, a person, a group) [0.144, 0.051, 0.706, 0.865]", "B. talk to (e.g., self, a person, a group) [0.343, 0.027, 0.586, 0.888] -> walk [0.279, 0.079, 0.473, 1.0] -> carry/hold (an object) [0.279, 0.079, 0.473, 1.0]", "C. take a photo [0.153, 0.000, 0.430, 0.865] -> carry/hold (an object) [0.424, 0.006, 0.292, 1.000] -> walk [0.315, 0.095, 0.412, 0.824]", "D. carry/hold (an object) [0.315, 0.000, 0.543, 1.000] -> run/jog [0.502, 0.127, 0.506, 0.777] -> walk [0.269, 0.000, 0.622, 0.955]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.279, 0.079, 0.473, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1306, "options": [ "A. [0.108, 0.062, 0.957, 0.959]", "B. [0.894, 0.463, 1.0, 1.0]", "C. [0.625, 0.437, 0.972, 0.788]", "D. [0.392, 0.259, 1.0, 0.542]" ], "question": "What location in the frame is the person currently in the [0.019, 0.076, 0.859, 0.96] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1327, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1344, "options": [ "A. shoot [0.782, 0.367, 0.747, 0.658] -> bend/bow (at the waist) [0.463, 0.279, 0.631, 0.675] -> carry/hold (an object) [0.444, 0.319, 0.740, 0.948]", "B. bend/bow (at the waist) [0.617, 0.204, 0.769, 0.795] -> carry/hold (an object) [0.617, 0.204, 0.769, 0.795] -> give/serve (an object) to (a person) [0.617, 0.204, 0.769, 0.795]", "C. hug (a person) [0.662, 0.120, 0.923, 0.732] -> bend/bow (at the waist) [0.679, 0.190, 0.885, 0.905] -> give/serve (an object) to (a person) [0.704, 0.236, 0.733, 0.719]", "D. give/serve (an object) to (a person) [0.647, 0.081, 0.855, 0.880] -> bend/bow (at the waist) [0.685, 0.138, 0.927, 0.890] -> play with pets [0.427, 0.109, 0.702, 0.867]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.617, 0.204, 0.769, 0.795] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1346, "options": [ "A. [0.203, 0.034, 0.54, 0.594]", "B. [0.68, 0.1, 0.881, 0.672]", "C. [0.483, 0.273, 0.751, 0.802]", "D. [0.024, 0.17, 0.876, 0.983]" ], "question": "Where is the person currently performing the bend/bow (at the waist) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1352, "options": [ "A. drink -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)", "C. drink -> kiss (a person)", "D. fight/hit (a person) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1500, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1501, "options": [ "A. [0.328, 0.262, 0.884, 0.963]", "B. [0.174, 0.119, 0.669, 0.893]", "C. [0.493, 0.35, 1.0, 0.747]", "D. [0.101, 0.388, 0.767, 1.0]" ], "question": "Where was the person currently performing the bend/bow (at the waist) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1630, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. sit, drink, listen to (a person), watch (a person)", "D. walk, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.001, 0.021, 0.389, 0.96] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1674, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. sit, carry/hold (an object), talk to (e.g., self, a person, a group)", "D. sit, watch (a person)" ], "question": "What action is the person at the location [0.0, 0.009, 0.502, 0.934] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1691, "options": [ "A. carry/hold (an object) -> crouch/kneel", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. fishing -> carry/hold (an object)", "D. work on a computer -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1738, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.828, 0.419, 0.964, 0.636] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1738 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 904, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.219, 0.091, 0.628, 0.893] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 917, "options": [ "A. [0.257, 0.095, 0.748, 0.884]", "B. [0.512, 0.099, 0.993, 0.906]", "C. [0.059, 0.0, 0.9, 0.626]", "D. [0.054, 0.28, 0.917, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 941, "options": [ "A. sit, smoke, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. sit, smoke, listen to (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.038, 0.099, 0.574, 0.931] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 986, "options": [ "A. [0.404, 0.335, 0.693, 0.984]", "B. [0.488, 0.102, 0.943, 0.911]", "C. [0.216, 0.098, 0.656, 0.891]", "D. [0.715, 0.174, 1.0, 0.666]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1076, "options": [ "A. [0.184, 0.369, 1.0, 1.0]", "B. [0.574, 0.0, 1.0, 0.897]", "C. [0.337, 0.469, 0.382, 0.54]", "D. [0.414, 0.098, 0.811, 0.924]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1113, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1148, "options": [ "A. sit, smoke, listen to (a person)", "B. stand, kiss (a person)", "C. sit, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.164, 0.088, 0.518, 0.895] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1213, "options": [ "A. [0.182, 0.19, 0.44, 0.899]", "B. [0.132, 0.0, 0.418, 0.72]", "C. [0.532, 0.097, 0.997, 0.899]", "D. [0.0, 0.468, 0.628, 1.0]" ], "question": "Where was the person currently performing the open (e.g., a window, a car door) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1247, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, listen to (a person)", "C. sit, talk to (e.g., self, a person, a group), watch (a person)", "D. walk, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.165, 0.178, 0.598, 0.898] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1341, "options": [ "A. [0.036, 0.07, 0.64, 0.879]", "B. [0.001, 0.074, 0.356, 0.929]", "C. [0.0, 0.233, 0.112, 1.0]", "D. [0.166, 0.298, 0.366, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1351, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1455, "options": [ "A. [0.06, 0.123, 0.518, 0.904]", "B. [0.338, 0.0, 0.943, 0.986]", "C. [0.05, 0.088, 0.915, 0.89]", "D. [0.102, 0.292, 0.629, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1482, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.324, 0.184, 0.646, 0.9] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1573, "options": [ "A. kick (a person)", "B. carry/hold (an object)", "C. walk and talk to (e.g., self, a person, a group)", "D. text on/look at a cellphone" ], "question": "What action is the person currently in the [0.059, 0.139, 0.471, 0.925] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1653, "options": [ "A. [0.0, 0.039, 0.354, 0.878]", "B. [0.318, 0.125, 0.923, 0.912]", "C. [0.117, 0.079, 0.645, 0.898]", "D. [0.221, 0.335, 0.768, 0.879]" ], "question": "What location in the frame is the person currently in the [0.542, 0.085, 1.0, 0.901] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1733, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1751, "options": [ "A. [0.0, 0.167, 0.276, 1.0]", "B. [0.06, 0.181, 0.855, 0.875]", "C. [0.048, 0.084, 0.546, 0.926]", "D. [0.223, 0.0, 0.276, 0.637]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1751 ]

[ { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 919, "options": [ "A. eat [0.627, 0.231, 0.894, 0.888] -> touch (an object) [0.627, 0.231, 0.894, 0.888] -> carry/hold (an object) [0.601, 0.246, 0.893, 0.905]", "B. open (e.g., a window, a car door) [0.672, 0.240, 1.000, 0.937] -> eat [0.590, 0.171, 1.000, 0.978] -> carry/hold (an object) [0.526, 0.362, 0.995, 0.969]", "C. open (e.g., a window, a car door) [0.534, 0.081, 0.856, 0.739] -> eat [0.472, 0.315, 1.000, 1.000] -> carry/hold (an object) [0.799, 0.444, 0.980, 0.830]", "D. talk to (e.g., self, a person, a group) [0.779, 0.127, 0.698, 0.997] -> eat [0.529, 0.355, 1.000, 0.802] -> carry/hold (an object) [0.554, 0.256, 0.778, 0.997]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.601, 0.246, 0.893, 0.905] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 940, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.537, 0.227, 0.889, 0.92] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 943, "options": [ "A. 5", "B. 3", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: touch (an object)?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 958, "options": [ "A. [0.465, 0.458, 0.711, 1.0]", "B. [0.14, 0.114, 0.444, 0.828]", "C. [0.6, 0.216, 0.89, 0.885]", "D. [0.533, 0.079, 1.0, 0.608]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1001, "options": [ "A. touch (an object) [0.294, 0.418, 0.726, 0.675] -> talk to (e.g., self, a person, a group) [0.446, 0.165, 0.596, 0.574] -> close (e.g., a door, a box) [0.388, 0.157, 0.544, 0.538]", "B. hug (a person) [0.498, 0.388, 0.823, 0.503] -> touch (an object) [0.388, 0.065, 0.685, 0.518] -> talk to (e.g., self, a person, a group) [0.617, 0.080, 0.868, 0.835]", "C. read [0.473, 0.062, 0.784, 0.567] -> eat [0.381, 0.428, 0.681, 0.611] -> talk to (e.g., self, a person, a group) [0.250, 0.293, 0.590, 0.716]", "D. touch (an object) [0.427, 0.228, 0.734, 0.666] -> talk to (e.g., self, a person, a group) [0.427, 0.228, 0.734, 0.666] -> eat [0.512, 0.246, 0.72, 0.632]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.512, 0.246, 0.72, 0.632] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1028, "options": [ "A. carry/hold (an object) [0.264, 0.233, 0.431, 0.655] -> eat [0.264, 0.233, 0.431, 0.655] -> touch (an object) [0.264, 0.233, 0.431, 0.655]", "B. touch (an object) [0.232, 0.204, 0.281, 0.530] -> touch (an object) [0.260, 0.047, 0.257, 0.805] -> carry/hold (an object) [0.259, 0.052, 0.413, 0.498]", "C. answer phone [0.264, 0.317, 0.503, 0.836] -> carry/hold (an object) [0.364, 0.429, 0.623, 0.782] -> eat [0.095, 0.193, 0.583, 0.721]", "D. touch (an object) [0.116, 0.227, 0.515, 0.516] -> carry/hold (an object) [0.342, 0.065, 0.490, 0.797] -> extract [0.261, 0.147, 0.476, 0.614]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.264, 0.233, 0.431, 0.655] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1037, "options": [ "A. [0.513, 0.267, 0.716, 0.655]", "B. [0.228, 0.049, 0.679, 0.586]", "C. [0.372, 0.52, 0.978, 0.75]", "D. [0.255, 0.253, 0.488, 0.672]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1057, "options": [ "A. eat [0.444, 0.163, 0.591, 0.576] -> take a photo [0.499, 0.179, 0.640, 0.707] -> talk to (e.g., self, a person, a group) [0.746, 0.165, 0.601, 0.610]", "B. dig [0.561, 0.303, 0.605, 0.731] -> eat [0.589, 0.295, 0.733, 0.571] -> touch (an object) [0.515, 0.184, 0.905, 0.792]", "C. talk to (e.g., self, a person, a group) [0.746, 0.143, 0.898, 0.751] -> smoke [0.499, 0.069, 0.934, 0.692] -> touch (an object) [0.709, 0.120, 0.689, 0.698]", "D. touch (an object) [0.569, 0.238, 0.751, 0.681] -> talk to (e.g., self, a person, a group) [0.569, 0.238, 0.751, 0.681] -> eat [0.5, 0.239, 0.756, 0.679]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.5, 0.239, 0.756, 0.679] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1062, "options": [ "A. [0.586, 0.06, 0.945, 0.519]", "B. [0.16, 0.158, 0.703, 0.878]", "C. [0.578, 0.542, 0.617, 0.69]", "D. [0.501, 0.26, 0.73, 0.651]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1110, "options": [ "A. touch (an object) [0.630, 0.360, 0.951, 0.794] -> jump/leap [0.490, 0.154, 0.737, 0.853] -> drink [0.726, 0.099, 0.613, 0.674]", "B. talk to (e.g., self, a person, a group) [0.600, 0.435, 0.942, 0.740] -> drink [0.561, 0.445, 0.917, 0.468] -> throw [0.754, 0.272, 0.684, 0.617]", "C. drink [0.59, 0.288, 0.771, 0.654] -> touch (an object) [0.59, 0.288, 0.771, 0.654] -> talk to (e.g., self, a person, a group) [0.59, 0.288, 0.771, 0.654]", "D. touch (an object) [0.724, 0.468, 0.897, 0.809] -> text on/look at a cellphone [0.709, 0.418, 0.603, 0.635] -> talk to (e.g., self, a person, a group) [0.782, 0.238, 0.953, 0.558]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.59, 0.288, 0.771, 0.654] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1120, "options": [ "A. [0.0, 0.0, 0.384, 0.61]", "B. [0.566, 0.285, 0.721, 0.664]", "C. [0.112, 0.19, 0.375, 0.878]", "D. [0.378, 0.0, 0.594, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1145, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, touch (an object), listen to (a person), watch (a person)", "C. sit, listen to (a person), watch (a person)", "D. sit, touch (an object)" ], "question": "What action is the person at the location [0.118, 0.115, 0.504, 0.91] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1176, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1206, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1289, "options": [ "A. [0.615, 0.255, 1.0, 0.961]", "B. [0.185, 0.125, 0.694, 0.857]", "C. [0.332, 0.167, 0.874, 0.897]", "D. [0.388, 0.0, 0.662, 0.734]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1306, "options": [ "A. open (e.g., a window, a car door)", "B. drink", "C. take a photo", "D. eat and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.142, 0.103, 0.679, 0.897] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1402, "options": [ "A. [0.62, 0.0, 0.841, 0.643]", "B. [0.502, 0.214, 0.623, 0.879]", "C. [0.558, 0.0, 0.39, 0.781]", "D. [0.594, 0.176, 0.888, 0.876]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1467, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.13, 0.187, 0.736, 0.903] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1553, "options": [ "A. 5", "B. 3", "C. 2", "D. 4" ], "question": "How many people in the current frame are performing the action: touch (an object)?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1658, "options": [ "A. [0.556, 0.403, 0.58, 0.957]", "B. [0.128, 0.118, 0.682, 0.897]", "C. [0.562, 0.0, 0.873, 0.624]", "D. [0.308, 0.142, 0.818, 0.878]" ], "question": "Where is the person currently performing the touch (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1672, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.568, 0.157, 0.884, 0.84] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1672 ]

[ { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1120, "options": [ "A. stand, watch (a person)", "B. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, write, listen to (a person)" ], "question": "What action is the person at the location [0.058, 0.353, 0.23, 0.617] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1148, "options": [ "A. 2", "B. 7", "C. 4", "D. 1" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1162, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.477, 0.205, 0.825, 0.979] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1170, "options": [ "A. walk [0.433, 0.119, 1.000, 0.925] -> sail boat [0.428, 0.291, 1.000, 0.967] -> carry/hold (an object) [0.362, 0.398, 0.885, 0.935]", "B. lie/sleep [0.624, 0.322, 1.000, 0.797] -> carry/hold (an object) [0.697, 0.216, 1.000, 0.869] -> talk to (e.g., self, a person, a group) [0.704, 0.308, 0.763, 0.865]", "C. walk [0.515, 0.247, 0.954, 0.981] -> carry/hold (an object) [0.515, 0.247, 0.954, 0.981] -> talk to (e.g., self, a person, a group) [0.515, 0.247, 0.954, 0.981]", "D. carry/hold (an object) [0.689, 0.094, 0.849, 0.933] -> extract [0.335, 0.175, 0.981, 0.957] -> talk to (e.g., self, a person, a group) [0.708, 0.255, 1.000, 0.797]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.515, 0.247, 0.954, 0.981] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1174, "options": [ "A. talk to (e.g., self, a person, a group) [0.5, 0.255, 0.971, 0.979] -> walk [0.549, 0.221, 0.993, 0.989] -> carry/hold (an object) [0.549, 0.221, 0.993, 0.989]", "B. carry/hold (an object) [0.492, 0.082, 1.000, 1.000] -> lie/sleep [0.695, 0.112, 1.000, 0.798] -> walk [0.375, 0.124, 1.000, 1.000]", "C. talk to (e.g., self, a person, a group) [0.415, 0.331, 0.824, 1.000] -> lie/sleep [0.625, 0.330, 1.000, 1.000] -> carry/hold (an object) [0.707, 0.169, 0.887, 0.796]", "D. talk to (e.g., self, a person, a group) [0.369, 0.197, 1.000, 0.931] -> walk [0.700, 0.375, 1.000, 0.896] -> carry/hold (an object) [0.470, 0.122, 0.973, 0.960]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.549, 0.221, 0.993, 0.989] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1197, "options": [ "A. [0.079, 0.0, 0.875, 0.996]", "B. [0.332, 0.02, 0.845, 0.706]", "C. [0.322, 0.105, 0.987, 0.997]", "D. [0.0, 0.0, 0.641, 1.0]" ], "question": "Where is the person currently performing the get up located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1264, "options": [ "A. 8", "B. 4", "C. 6", "D. 3" ], "question": "How many people were performing the listen to (a person) in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1316, "options": [ "A. talk to (e.g., self, a person, a group)", "B. answer phone and lie/sleep", "C. smoke", "D. fall down" ], "question": "What action is the person currently in the [0.271, 0.067, 1.0, 0.993] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1396, "options": [ "A. [0.626, 0.305, 0.894, 1.0]", "B. [0.429, 0.087, 0.908, 0.974]", "C. [0.721, 0.0, 0.968, 0.818]", "D. [0.091, 0.185, 0.347, 0.992]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1433, "options": [ "A. 3", "B. 4", "C. 8", "D. 5" ], "question": "How many people were performing the sit in the scene 300 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1518, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.493, 0.161, 0.957, 0.973] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1522, "options": [ "A. talk to (e.g., self, a person, a group) [0.512, 0.181, 0.889, 0.955] -> walk [0.411, 0.144, 0.798, 0.97] -> carry/hold (an object) [0.411, 0.144, 0.798, 0.97]", "B. fall down [0.437, 0.221, 0.825, 0.953] -> walk [0.371, 0.286, 0.740, 1.000] -> talk to (e.g., self, a person, a group) [0.589, 0.343, 0.878, 0.776]", "C. carry/hold (an object) [0.558, 0.108, 0.972, 0.824] -> talk to (e.g., self, a person, a group) [0.322, 0.211, 0.734, 1.000] -> text on/look at a cellphone [0.383, 0.170, 0.772, 0.827]", "D. carry/hold (an object) [0.502, 0.305, 0.978, 0.922] -> talk to (e.g., self, a person, a group) [0.506, 0.271, 0.949, 0.994] -> eat [0.223, 0.335, 0.954, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.411, 0.144, 0.798, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1551, "options": [ "A. 4", "B. 2", "C. 1", "D. 3" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1599, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.385, 0.313, 0.592, 0.692] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1625, "options": [ "A. 4", "B. 5", "C. 2", "D. 3" ], "question": "How many people in the current frame are performing the action: touch (an object)?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1714, "options": [ "A. 7", "B. 6", "C. 4", "D. 3" ], "question": "How many people were performing the listen to (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1743, "options": [ "A. talk to (e.g., self, a person, a group) [0.320, 0.180, 0.967, 0.686] -> play board game [0.527, 0.216, 1.000, 0.833] -> carry/hold (an object) [0.348, 0.487, 0.646, 0.740]", "B. talk to (e.g., self, a person, a group) [0.349, 0.341, 0.826, 0.833] -> carry/hold (an object) [0.407, 0.344, 0.818, 0.837] -> touch (an object) [0.407, 0.344, 0.818, 0.837]", "C. talk to (e.g., self, a person, a group) [0.448, 0.529, 0.877, 0.665] -> carry/hold (an object) [0.241, 0.353, 0.965, 0.753] -> work on a computer [0.544, 0.517, 0.724, 0.683]", "D. touch (an object) [0.435, 0.423, 0.674, 0.707] -> kick (an object) [0.258, 0.315, 0.869, 0.997] -> carry/hold (an object) [0.272, 0.191, 0.775, 0.851]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.407, 0.344, 0.818, 0.837] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1745, "options": [ "A. stand, carry/hold (an object), listen to (a person), watch (a person)", "B. stand, write", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.01, 0.241, 0.095, 0.621] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1765, "options": [ "A. carry/hold (an object) [0.366, 0.349, 0.811, 0.81] -> touch (an object) [0.366, 0.349, 0.811, 0.81] -> talk to (e.g., self, a person, a group) [0.366, 0.349, 0.811, 0.81]", "B. swim [0.448, 0.346, 0.979, 0.807] -> carry/hold (an object) [0.461, 0.234, 0.758, 0.745] -> talk to (e.g., self, a person, a group) [0.401, 0.391, 0.816, 0.702]", "C. talk to (e.g., self, a person, a group) [0.510, 0.162, 0.650, 0.766] -> carry/hold (an object) [0.313, 0.275, 0.760, 0.953] -> hug (a person) [0.498, 0.198, 0.661, 0.969]", "D. touch (an object) [0.338, 0.405, 0.808, 0.659] -> talk to (e.g., self, a person, a group) [0.205, 0.327, 0.974, 0.971] -> fishing [0.316, 0.447, 0.663, 0.797]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.366, 0.349, 0.811, 0.81] location?", "sub_answer_type": "Action Trajectory" } ]

[ 1000, 1765 ]

[ { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 955, "options": [ "A. read -> talk to (e.g., self, a person, a group)", "B. answer phone -> read", "C. talk to (e.g., self, a person, a group) -> crawl", "D. read -> point to (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 994, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1056, "options": [ "A. [0.252, 0.0, 0.786, 0.816]", "B. [0.006, 0.0, 0.401, 0.702]", "C. [0.052, 0.2, 0.382, 1.0]", "D. [0.106, 0.09, 0.637, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1069, "options": [ "A. sit, listen to (a person), watch (a person)", "B. sit, play musical instrument", "C. stand, listen to (a person), watch (a person)", "D. stand, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.019, 0.013, 0.442, 0.993] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1151, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1220, "options": [ "A. stand, touch (an object), listen to (a person), watch (a person)", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. sit, watch (a person)" ], "question": "What action is the person at the location [0.0, 0.076, 0.251, 0.996] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1223, "options": [ "A. [0.657, 0.226, 0.975, 0.904]", "B. [0.278, 0.045, 0.979, 0.991]", "C. [0.49, 0.495, 0.808, 0.887]", "D. [0.8, 0.499, 1.0, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1240, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> get up", "B. touch (an object) -> talk to (e.g., self, a person, a group) -> carry/hold (an object)", "C. crouch/kneel -> touch (an object) -> carry/hold (an object)", "D. close (e.g., a door, a box) -> carry/hold (an object) -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1301, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, listen to (a person)", "C. sit, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.002, 0.04, 0.412, 0.987] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1361, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1379, "options": [ "A. [0.194, 0.116, 0.518, 1.0]", "B. [0.154, 0.025, 0.24, 0.658]", "C. [0.066, 0.24, 0.346, 0.857]", "D. [0.004, 0.481, 0.333, 0.934]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1402, "options": [ "A. [0.0, 0.344, 0.273, 1.0]", "B. [0.063, 0.198, 0.495, 0.98]", "C. [0.193, 0.606, 0.338, 0.962]", "D. [0.321, 0.276, 0.551, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1426, "options": [ "A. carry/hold (an object) -> take a photo", "B. exit -> carry/hold (an object)", "C. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "D. carry/hold (an object) -> run/jog" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1491, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1597, "options": [ "A. sit, listen to (a person)", "B. stand, touch (an object)", "C. stand, listen to (a person), watch (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.014, 0.225, 0.493, 0.959] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1622, "options": [ "A. [0.3, 0.014, 0.506, 0.825]", "B. [0.455, 0.571, 0.544, 0.891]", "C. [0.514, 0.3, 0.74, 0.693]", "D. [0.71, 0.258, 0.98, 0.907]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1624, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1684, "options": [ "A. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "B. bend/bow (at the waist) -> walk", "C. talk to (e.g., self, a person, a group) -> watch (e.g., TV)", "D. play with pets -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1704, "options": [ "A. paint", "B. take a photo", "C. talk to (e.g., self, a person, a group)", "D. carry/hold (an object) and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.174, 0.155, 0.47, 0.99] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1737, "options": [ "A. carry/hold (an object) -> hit (an object) -> talk to (e.g., self, a person, a group)", "B. hit (an object) -> talk to (e.g., self, a person, a group) -> touch (an object)", "C. talk to (e.g., self, a person, a group) -> touch (an object) -> shoot", "D. bend/bow (at the waist) -> talk to (e.g., self, a person, a group) -> hit (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1773, "options": [ "A. [0.063, 0.187, 0.536, 0.985]", "B. [0.102, 0.122, 0.6, 0.712]", "C. [0.008, 0.482, 0.827, 1.0]", "D. [0.184, 0.173, 0.895, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1773 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 963, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.029, 0.103, 0.576, 0.943] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 977, "options": [ "A. walk -> carry/hold (an object)", "B. get up -> walk", "C. walk -> climb (e.g., a mountain)", "D. crouch/kneel -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1029, "options": [ "A. eat -> carry/hold (an object) -> drink", "B. crouch/kneel -> carry/hold (an object) -> drink", "C. text on/look at a cellphone -> carry/hold (an object) -> crouch/kneel", "D. carry/hold (an object) -> crouch/kneel -> sing to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1050, "options": [ "A. [0.0, 0.0, 0.451, 0.801]", "B. [0.414, 0.073, 0.96, 0.991]", "C. [0.0, 0.295, 0.44, 1.0]", "D. [0.051, 0.058, 0.623, 1.0]" ], "question": "Where is the person currently performing the play musical instrument located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1051, "options": [ "A. [0.056, 0.002, 0.574, 0.99]", "B. [0.216, 0.604, 0.351, 0.992]", "C. [0.227, 0.119, 0.742, 1.0]", "D. [0.0, 0.163, 0.318, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1142, "options": [ "A. [0.518, 0.018, 0.999, 0.95]", "B. [0.326, 0.29, 1.0, 1.0]", "C. [0.454, 0.0, 0.708, 0.784]", "D. [0.4, 0.609, 0.516, 0.958]" ], "question": "Where is the person currently performing the play musical instrument located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1161, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1213, "options": [ "A. fishing [0.434, 0.319, 0.287, 0.615] -> talk to (e.g., self, a person, a group) [0.261, 0.439, 0.361, 0.719] -> give/serve (an object) to (a person) [0.286, 0.196, 0.574, 0.688]", "B. talk to (e.g., self, a person, a group) [0.328, 0.352, 0.476, 0.687] -> carry/hold (an object) [0.353, 0.343, 0.472, 0.678] -> give/serve (an object) to (a person) [0.353, 0.343, 0.472, 0.678]", "C. talk to (e.g., self, a person, a group) [0.494, 0.340, 0.291, 0.881] -> point to (an object) [0.165, 0.519, 0.369, 0.648] -> carry/hold (an object) [0.246, 0.527, 0.614, 0.659]", "D. give/serve (an object) to (a person) [0.329, 0.260, 0.319, 0.648] -> ride (e.g., a bike, a car, a horse) [0.457, 0.493, 0.559, 0.602] -> carry/hold (an object) [0.495, 0.354, 0.628, 0.676]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.353, 0.343, 0.472, 0.678] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1215, "options": [ "A. talk to (e.g., self, a person, a group) [0.154, 0.460, 0.334, 0.661] -> carry/hold (an object) [0.380, 0.172, 0.646, 0.787] -> bend/bow (at the waist) [0.242, 0.515, 0.532, 0.677]", "B. bend/bow (at the waist) [0.325, 0.338, 0.484, 0.715] -> carry/hold (an object) [0.325, 0.338, 0.484, 0.715] -> talk to (e.g., self, a person, a group) [0.325, 0.338, 0.484, 0.715]", "C. row boat [0.380, 0.363, 0.363, 0.768] -> bend/bow (at the waist) [0.384, 0.496, 0.609, 0.574] -> talk to (e.g., self, a person, a group) [0.382, 0.445, 0.613, 0.627]", "D. talk to (e.g., self, a person, a group) [0.424, 0.341, 0.620, 0.880] -> brush teeth [0.430, 0.439, 0.373, 0.516] -> carry/hold (an object) [0.270, 0.276, 0.648, 0.873]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.325, 0.338, 0.484, 0.715] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1269, "options": [ "A. walk, listen to (a person)", "B. sit, play musical instrument, watch (a person)", "C. carry/hold (an object), crouch/kneel", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.338, 0.434, 0.548, 0.991] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1276, "options": [ "A. [0.514, 0.512, 0.79, 0.999]", "B. [0.267, 0.623, 0.656, 0.915]", "C. [0.456, 0.332, 0.782, 0.78]", "D. [0.564, 0.0, 1.0, 0.98]" ], "question": "What location in the frame is the person currently in the [0.059, 0.48, 0.431, 0.992] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1288, "options": [ "A. [0.385, 0.059, 0.485, 0.696]", "B. [0.363, 0.364, 0.735, 0.814]", "C. [0.543, 0.059, 0.975, 0.99]", "D. [0.111, 0.291, 0.6, 0.986]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1353, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.041, 0.323, 0.218, 0.933] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1388, "options": [ "A. 4", "B. 5", "C. 3", "D. 7" ], "question": "How many people were performing the sit in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1442, "options": [ "A. sing to (e.g., self, a person, a group) -> carry/hold (an object)", "B. push (another person) -> play musical instrument", "C. carry/hold (an object) -> play musical instrument", "D. carry/hold (an object) -> shoot" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1492, "options": [ "A. smoke", "B. kiss (a person)", "C. carry/hold (an object) and talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.492, 0.214, 0.938, 0.988] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1511, "options": [ "A. 3", "B. 6", "C. 8", "D. 9" ], "question": "How many people were performing the watch (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1535, "options": [ "A. [0.316, 0.087, 0.962, 0.981]", "B. [0.117, 0.034, 0.864, 0.977]", "C. [0.19, 0.319, 0.87, 1.0]", "D. [0.0, 0.0, 0.782, 0.693]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1557, "options": [ "A. open (e.g., a window, a car door) [0.275, 0.360, 0.385, 0.757] -> grab (a person) [0.105, 0.370, 0.228, 0.813] -> talk to (e.g., self, a person, a group) [0.244, 0.515, 0.312, 0.548] -> walk [0.167, 0.355, 0.141, 0.580]", "B. walk [0.468, 0.285, 0.534, 0.587] -> carry/hold (an object) [0.156, 0.564, 0.460, 0.555] -> talk to (e.g., self, a person, a group) [0.279, 0.622, 0.457, 0.730] -> play with kids [0.470, 0.384, 0.523, 0.617]", "C. grab (a person) [0.122, 0.419, 0.194, 0.838] -> play board game [0.229, 0.286, 0.178, 0.775] -> carry/hold (an object) [0.247, 0.496, 0.148, 0.882] -> talk to (e.g., self, a person, a group) [0.211, 0.544, 0.526, 0.852]", "D. walk [0.279, 0.469, 0.34, 0.748] -> carry/hold (an object) [0.279, 0.469, 0.34, 0.748] -> grab (a person) [0.279, 0.469, 0.34, 0.748] -> talk to (e.g., self, a person, a group) [0.279, 0.469, 0.34, 0.748]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.279, 0.469, 0.34, 0.748] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1558, "options": [ "A. carry/hold (an object) [0.103, 0.514, 0.209, 0.918] -> talk to (e.g., self, a person, a group) [0.470, 0.476, 0.372, 0.571] -> grab (a person) [0.178, 0.415, 0.211, 0.698] -> watch (e.g., TV) [0.167, 0.270, 0.431, 0.570]", "B. walk [0.279, 0.469, 0.34, 0.748] -> carry/hold (an object) [0.279, 0.469, 0.34, 0.748] -> grab (a person) [0.279, 0.469, 0.34, 0.748] -> talk to (e.g., self, a person, a group) [0.279, 0.469, 0.34, 0.748]", "C. walk [0.459, 0.669, 0.176, 0.703] -> open (e.g., a window, a car door) [0.370, 0.657, 0.368, 0.922] -> talk to (e.g., self, a person, a group) [0.446, 0.468, 0.404, 0.715] -> carry/hold (an object) [0.458, 0.462, 0.381, 0.655]", "D. touch (an object) [0.393, 0.395, 0.313, 0.636] -> walk [0.408, 0.574, 0.175, 0.914] -> talk to (e.g., self, a person, a group) [0.474, 0.598, 0.174, 0.590] -> grab (a person) [0.396, 0.387, 0.478, 0.683]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.279, 0.469, 0.34, 0.748] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1576, "options": [ "A. carry/hold (an object) -> take (an object) from (a person)", "B. answer phone -> carry/hold (an object)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> swim" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1606, "options": [ "A. 7", "B. 4", "C. 3", "D. 6" ], "question": "How many people were performing the watch (a person) in the scene 300 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1626, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.645, 0.453, 0.758, 0.86] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1699, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.717, 0.472, 0.887, 0.778] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1699, "options": [ "A. talk to (e.g., self, a person, a group) [0.642, 0.635, 0.955, 0.715] -> shovel [0.648, 0.339, 0.893, 0.666] -> bend/bow (at the waist) [0.852, 0.523, 0.800, 0.672]", "B. shovel [0.566, 0.500, 0.741, 0.590] -> bend/bow (at the waist) [0.567, 0.315, 0.918, 0.627] -> catch (an object) [0.708, 0.495, 0.848, 0.834]", "C. bend/bow (at the waist) [0.717, 0.472, 0.887, 0.778] -> dig [0.717, 0.472, 0.887, 0.778] -> shovel [0.717, 0.472, 0.887, 0.778]", "D. shovel [0.712, 0.632, 0.758, 0.929] -> hand shake [0.678, 0.357, 0.883, 0.672] -> dig [0.623, 0.650, 0.831, 0.675]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.717, 0.472, 0.887, 0.778] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1709, "options": [ "A. lift/pick up -> turn (e.g., a screwdriver) -> swim", "B. carry/hold (an object) -> lift/pick up -> turn (e.g., a screwdriver)", "C. lift/pick up -> carry/hold (an object) -> write", "D. lift/pick up -> carry/hold (an object) -> dress/put on clothing" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1744, "options": [ "A. [0.317, 0.0, 0.57, 0.953]", "B. [0.523, 0.562, 0.57, 0.734]", "C. [0.815, 0.108, 1.0, 1.0]", "D. [0.583, 0.022, 0.785, 0.934]" ], "question": "Where is the person currently performing the open (e.g., a window, a car door) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1744 ]

[ { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 903, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 910, "options": [ "A. [0.138, 0.185, 0.423, 0.956]", "B. [0.021, 0.278, 0.894, 1.0]", "C. [0.219, 0.013, 1.0, 0.865]", "D. [0.026, 0.0, 0.809, 0.975]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 949, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. write -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> bend/bow (at the waist)", "D. carry/hold (an object) -> cut" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 956, "options": [ "A. walk [0.185, 0.136, 0.546, 0.968] -> carry/hold (an object) [0.185, 0.136, 0.546, 0.968] -> talk to (e.g., self, a person, a group) [0.185, 0.136, 0.546, 0.968]", "B. walk [0.170, 0.276, 0.742, 1.000] -> talk to (e.g., self, a person, a group) [0.122, 0.278, 0.687, 0.839] -> dig [0.329, 0.288, 0.646, 0.941]", "C. talk to (e.g., self, a person, a group) [0.000, 0.188, 0.690, 0.923] -> carry/hold (an object) [0.207, 0.029, 0.497, 1.000] -> lift/pick up [0.179, 0.289, 0.441, 1.000]", "D. crawl [0.273, 0.142, 0.620, 0.874] -> carry/hold (an object) [0.169, 0.000, 0.513, 0.931] -> talk to (e.g., self, a person, a group) [0.141, 0.110, 0.401, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.185, 0.136, 0.546, 0.968] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1025, "options": [ "A. hug (a person) [0.396, 0.097, 0.491, 0.751] -> carry/hold (an object) [0.367, 0.365, 0.827, 0.736] -> talk to (e.g., self, a person, a group) [0.285, 0.091, 0.643, 0.757]", "B. extract [0.360, 0.232, 0.667, 0.702] -> carry/hold (an object) [0.428, 0.063, 0.536, 0.782] -> talk to (e.g., self, a person, a group) [0.550, 0.057, 0.462, 0.702]", "C. carry/hold (an object) [0.262, 0.213, 0.606, 0.671] -> walk [0.436, 0.319, 0.815, 0.751] -> push (an object) [0.258, 0.067, 0.747, 0.625]", "D. walk [0.447, 0.232, 0.659, 0.819] -> carry/hold (an object) [0.447, 0.232, 0.659, 0.819] -> talk to (e.g., self, a person, a group) [0.447, 0.232, 0.659, 0.819]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.447, 0.232, 0.659, 0.819] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1029, "options": [ "A. stand, carry/hold (an object), listen to (a person), watch (a person)", "B. crouch/kneel, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "D. walk, carry/hold (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.126, 0.281, 0.333, 0.953] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1186, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.556, 0.059, 0.887, 0.99] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1193, "options": [ "A. lie/sleep -> kick (a person)", "B. play musical instrument -> lie/sleep", "C. talk to (e.g., self, a person, a group) -> run/jog", "D. talk to (e.g., self, a person, a group) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1204, "options": [ "A. [0.201, 0.416, 0.419, 0.994]", "B. [0.386, 0.066, 1.0, 0.739]", "C. [0.385, 0.246, 1.0, 1.0]", "D. [0.111, 0.056, 0.924, 0.945]" ], "question": "Where is the person currently performing the lie/sleep located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1240, "options": [ "A. hand wave [0.121, 0.097, 0.402, 0.930] -> carry/hold (an object) [0.381, 0.128, 0.550, 1.000] -> talk to (e.g., self, a person, a group) [0.124, 0.006, 0.339, 0.774]", "B. fall down [0.367, 0.051, 0.444, 0.999] -> carry/hold (an object) [0.186, 0.134, 0.597, 1.000] -> walk [0.104, 0.000, 0.347, 1.000]", "C. walk [0.227, 0.042, 0.518, 0.974] -> carry/hold (an object) [0.227, 0.042, 0.518, 0.974] -> talk to (e.g., self, a person, a group) [0.227, 0.042, 0.518, 0.974]", "D. get up [0.410, 0.000, 0.443, 1.000] -> talk to (e.g., self, a person, a group) [0.303, 0.070, 0.371, 0.960] -> carry/hold (an object) [0.317, 0.000, 0.712, 0.915]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.227, 0.042, 0.518, 0.974] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1249, "options": [ "A. talk to (e.g., self, a person, a group) [0.297, 0.000, 0.683, 0.981] -> fight/hit (a person) [0.614, 0.180, 0.791, 0.913] -> carry/hold (an object) [0.669, 0.079, 1.000, 0.806]", "B. talk to (e.g., self, a person, a group) [0.637, 0.005, 0.751, 0.962] -> work on a computer [0.779, 0.000, 0.913, 0.935] -> carry/hold (an object) [0.638, 0.000, 0.836, 1.000]", "C. carry/hold (an object) [0.545, 0.031, 1.000, 1.000] -> martial art [0.571, 0.000, 0.963, 0.972] -> walk [0.579, 0.191, 0.824, 0.957]", "D. talk to (e.g., self, a person, a group) [0.496, 0.041, 0.863, 0.986] -> walk [0.641, 0.0, 0.984, 0.983] -> carry/hold (an object) [0.641, 0.0, 0.984, 0.983]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.641, 0.0, 0.984, 0.983] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1250, "options": [ "A. [0.202, 0.0, 0.401, 1.0]", "B. [0.25, 0.142, 0.668, 0.993]", "C. [0.479, 0.056, 0.832, 0.947]", "D. [0.259, 0.439, 0.706, 1.0]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1257, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> text on/look at a cellphone", "B. take a photo -> walk -> carry/hold (an object)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> walk", "D. run/jog -> talk to (e.g., self, a person, a group) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1266, "options": [ "A. [0.642, 0.135, 0.996, 1.0]", "B. [0.004, 0.48, 0.872, 0.995]", "C. [0.382, 0.039, 0.827, 0.99]", "D. [0.204, 0.054, 0.572, 1.0]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1365, "options": [ "A. carry/hold (an object) -> touch (an object)", "B. touch (an object) -> play with kids", "C. crouch/kneel -> carry/hold (an object)", "D. touch (an object) -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1414, "options": [ "A. [0.159, 0.186, 0.427, 0.923]", "B. [0.494, 0.112, 0.649, 0.977]", "C. [0.339, 0.017, 0.378, 0.703]", "D. [0.793, 0.169, 0.66, 0.809]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1434, "options": [ "A. [0.017, 0.203, 0.256, 0.951]", "B. [0.0, 0.0, 0.0, 1.0]", "C. [0.197, 0.047, 0.976, 0.994]", "D. [0.0, 0.378, 0.0, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1559, "options": [ "A. walk and carry/hold (an object)", "B. carry/hold (an object) and touch (an object)", "C. fall down", "D. martial art" ], "question": "What action is the person currently in the [0.164, 0.329, 0.375, 0.994] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1561, "options": [ "A. talk to (e.g., self, a person, a group) [0.164, 0.329, 0.375, 0.994] -> carry/hold (an object) [0.166, 0.323, 0.376, 0.989] -> touch (an object) [0.166, 0.323, 0.376, 0.989]", "B. touch (an object) [0.163, 0.183, 0.569, 1.000] -> talk to (e.g., self, a person, a group) [0.327, 0.234, 0.305, 0.874] -> touch (an object) [0.143, 0.459, 0.462, 0.794]", "C. carry/hold (an object) [0.075, 0.401, 0.533, 1.000] -> fight/hit (a person) [0.144, 0.378, 0.302, 1.000] -> touch (an object) [0.364, 0.193, 0.286, 0.925]", "D. touch (an object) [0.259, 0.210, 0.553, 1.000] -> push (another person) [0.249, 0.184, 0.549, 1.000] -> talk to (e.g., self, a person, a group) [0.136, 0.357, 0.319, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.166, 0.323, 0.376, 0.989] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1562, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.057, 0.027, 0.901, 0.984] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1585, "options": [ "A. [0.055, 0.043, 0.688, 0.986]", "B. [0.027, 0.0, 0.551, 0.764]", "C. [0.52, 0.125, 0.907, 1.0]", "D. [0.274, 0.183, 0.781, 0.986]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1618, "options": [ "A. carry/hold (an object) -> lie/sleep", "B. stir -> lie/sleep", "C. pull (an object) -> lie/sleep", "D. lie/sleep -> text on/look at a cellphone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1703, "options": [ "A. [0.4, 0.259, 0.68, 0.887]", "B. [0.018, 0.009, 0.468, 0.983]", "C. [0.369, 0.528, 0.672, 1.0]", "D. [0.186, 0.54, 0.456, 0.986]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1726, "options": [ "A. [0.0, 0.315, 0.532, 1.0]", "B. [0.001, 0.428, 0.881, 0.99]", "C. [0.0, 0.038, 0.765, 1.0]", "D. [0.291, 0.0, 0.649, 0.719]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1769, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1769 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 917, "options": [ "A. [0.039, 0.0, 0.837, 0.986]", "B. [0.586, 0.0, 1.0, 1.0]", "C. [0.294, 0.0, 0.976, 0.974]", "D. [0.398, 0.286, 1.0, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 966, "options": [ "A. talk to (e.g., self, a person, a group) -> walk", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> dance", "D. talk to (e.g., self, a person, a group) -> shoot" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1008, "options": [ "A. martial art -> read", "B. lie/sleep -> read", "C. read -> talk to (e.g., self, a person, a group)", "D. exit -> read" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1031, "options": [ "A. talk to (e.g., self, a person, a group) -> read", "B. dance -> talk to (e.g., self, a person, a group)", "C. dig -> talk to (e.g., self, a person, a group)", "D. hand wave -> read" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1074, "options": [ "A. [0.591, 0.106, 0.841, 1.0]", "B. [0.349, 0.286, 0.708, 0.998]", "C. [0.002, 0.0, 0.586, 1.0]", "D. [0.088, 0.533, 0.423, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1083, "options": [ "A. [0.21, 0.094, 0.573, 0.798]", "B. [0.454, 0.041, 0.903, 1.0]", "C. [0.21, 0.233, 0.641, 0.989]", "D. [0.385, 0.646, 0.542, 0.985]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1102, "options": [ "A. [0.346, 0.056, 0.938, 0.976]", "B. [0.548, 0.0, 0.969, 0.691]", "C. [0.046, 0.012, 0.685, 1.0]", "D. [0.372, 0.0, 0.69, 0.798]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1232, "options": [ "A. [0.065, 0.042, 0.557, 0.983]", "B. [0.0, 0.063, 0.38, 0.71]", "C. [0.36, 0.028, 0.99, 0.993]", "D. [0.0, 0.337, 0.364, 0.859]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1330, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. paint -> talk to (e.g., self, a person, a group)", "C. put down -> carry/hold (an object)", "D. talk to (e.g., self, a person, a group) -> crawl" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1370, "options": [ "A. walk -> carry/hold (an object)", "B. carry/hold (an object) -> drink", "C. walk -> crawl", "D. lie/sleep -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1419, "options": [ "A. [0.351, 0.005, 0.755, 1.0]", "B. [0.631, 0.211, 0.969, 0.97]", "C. [0.361, 0.355, 0.463, 0.852]", "D. [0.367, 0.422, 1.0, 0.991]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1419, "options": [ "A. sit, talk to (e.g., self, a person, a group)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. stand, listen to (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.039, 0.037, 0.531, 0.842] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1435, "options": [ "A. close (e.g., a door, a box)", "B. push (another person)", "C. walk and carry/hold (an object)", "D. carry/hold (an object) and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.012, 0.06, 0.517, 0.909] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1543, "options": [ "A. [0.642, 0.212, 0.965, 0.995]", "B. [0.157, 0.345, 0.966, 0.976]", "C. [0.366, 0.098, 0.948, 0.932]", "D. [0.146, 0.0, 0.85, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1600, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, carry/hold (an object), touch (an object), talk to (e.g., self, a person, a group)", "C. stand, listen to (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.012, 0.024, 0.449, 0.985] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1691, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.757, 0.0, 1.0, 0.976] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1766, "options": [ "A. [0.002, 0.158, 0.157, 1.0]", "B. [0.448, 0.243, 1.0, 1.0]", "C. [0.368, 0.0, 0.975, 0.937]", "D. [0.074, 0.176, 1.0, 1.0]" ], "question": "What location in the frame is the person currently in the [0.195, 0.095, 0.523, 0.939] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1774, "options": [ "A. [0.195, 0.095, 0.523, 0.939]", "B. [0.046, 0.151, 0.311, 1.0]", "C. [0.31, 0.039, 0.788, 0.957]", "D. [0.412, 0.0, 0.308, 0.655]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1774 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 905, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.538, 0.212, 0.641, 0.568] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 933, "options": [ "A. walk and carry/hold (an object)", "B. point to (an object)", "C. cut", "D. bend/bow (at the waist)" ], "question": "What action is the person currently in the [0.386, 0.203, 0.577, 0.71] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 934, "options": [ "A. bend/bow (at the waist) [0.298, 0.037, 0.596, 0.873] -> dig [0.298, 0.037, 0.596, 0.873] -> shovel [0.298, 0.037, 0.596, 0.873]", "B. bend/bow (at the waist) [0.359, 0.216, 0.532, 0.876] -> smoke [0.128, 0.000, 0.640, 0.886] -> dig [0.246, 0.080, 0.437, 1.000]", "C. dig [0.355, 0.000, 0.656, 0.934] -> shovel [0.131, 0.228, 0.682, 0.778] -> lift/pick up [0.310, 0.192, 0.751, 0.911]", "D. shovel [0.197, 0.193, 0.438, 0.794] -> close (e.g., a door, a box) [0.245, 0.088, 0.487, 0.873] -> dig [0.430, 0.000, 0.430, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.298, 0.037, 0.596, 0.873] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 946, "options": [ "A. [0.286, 0.495, 0.744, 1.0]", "B. [0.314, 0.199, 0.497, 0.877]", "C. [0.48, 0.208, 0.657, 0.69]", "D. [0.328, 0.361, 0.364, 1.0]" ], "question": "Where was the person currently performing the climb (e.g., a mountain) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1003, "options": [ "A. walk -> eat", "B. kiss (a person) -> carry/hold (an object)", "C. carry/hold (an object) -> walk", "D. carry/hold (an object) -> take a photo" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1110, "options": [ "A. [0.095, 0.192, 0.473, 0.697]", "B. [0.34, 0.109, 0.526, 0.942]", "C. [0.543, 0.228, 0.607, 1.0]", "D. [0.608, 0.195, 0.766, 0.803]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1132, "options": [ "A. [0.174, 0.308, 0.711, 1.0]", "B. [0.078, 0.126, 0.564, 1.0]", "C. [0.054, 0.0, 1.0, 0.75]", "D. [0.104, 0.101, 0.997, 0.983]" ], "question": "Where was the person currently performing the lie/sleep in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1172, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.054, 0.969, 0.888] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1173, "options": [ "A. throw -> talk to (e.g., self, a person, a group)", "B. lie/sleep -> take a photo", "C. talk to (e.g., self, a person, a group) -> lie/sleep", "D. lie/sleep -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1219, "options": [ "A. carry/hold (an object) [0.322, 0.039, 0.808, 0.974] -> smoke [0.322, 0.039, 0.808, 0.974] -> talk to (e.g., self, a person, a group) [0.322, 0.039, 0.808, 0.974]", "B. talk to (e.g., self, a person, a group) [0.397, 0.203, 0.707, 0.960] -> climb (e.g., a mountain) [0.144, 0.000, 0.707, 0.951] -> smoke [0.189, 0.104, 0.852, 0.864]", "C. carry/hold (an object) [0.310, 0.000, 0.952, 0.951] -> crawl [0.301, 0.152, 0.989, 0.883] -> smoke [0.381, 0.058, 0.755, 0.969]", "D. talk to (e.g., self, a person, a group) [0.415, 0.001, 0.936, 0.842] -> carry/hold (an object) [0.233, 0.055, 0.828, 0.984] -> take (an object) from (a person) [0.465, 0.223, 0.933, 0.953]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.322, 0.039, 0.808, 0.974] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1238, "options": [ "A. [0.549, 0.262, 0.771, 0.624]", "B. [0.181, 0.487, 0.347, 0.803]", "C. [0.41, 0.2, 0.581, 0.941]", "D. [0.428, 0.417, 0.821, 1.0]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1241, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.38, 0.106, 0.59, 0.918] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1279, "options": [ "A. carry/hold (an object) -> turn (e.g., a screwdriver)", "B. bend/bow (at the waist) -> chop", "C. carry/hold (an object) -> watch (e.g., TV)", "D. carry/hold (an object) -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1373, "options": [ "A. kiss (a person) -> touch (an object)", "B. martial art -> touch (an object)", "C. turn (e.g., a screwdriver) -> touch (an object)", "D. touch (an object) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1398, "options": [ "A. [0.582, 0.299, 0.837, 1.0]", "B. [0.659, 0.0, 0.87, 0.732]", "C. [0.043, 0.13, 0.589, 0.998]", "D. [0.393, 0.144, 0.607, 0.977]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1421, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.208, 0.038, 0.776, 0.938] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1421, "options": [ "A. carry/hold (an object) [0.091, 0.102, 0.772, 0.795] -> touch (an object) [0.149, 0.170, 0.698, 0.814] -> push (another person) [0.159, 0.000, 0.636, 0.915]", "B. carry/hold (an object) [0.151, 0.033, 0.777, 0.933] -> read [0.151, 0.033, 0.777, 0.933] -> touch (an object) [0.151, 0.033, 0.777, 0.933]", "C. dress/put on clothing [0.308, 0.020, 0.965, 0.992] -> touch (an object) [0.171, 0.191, 0.834, 1.000] -> read [0.216, 0.169, 0.639, 1.000]", "D. carry/hold (an object) [0.205, 0.000, 0.741, 0.856] -> answer phone [0.086, 0.072, 0.734, 0.819] -> touch (an object) [0.201, 0.037, 0.950, 0.774]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.151, 0.033, 0.777, 0.933] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1449, "options": [ "A. [0.0, 0.0, 0.78, 0.875]", "B. [0.128, 0.023, 1.0, 0.707]", "C. [0.275, 0.359, 0.484, 0.847]", "D. [0.27, 0.024, 1.0, 0.999]" ], "question": "Where was the person currently performing the read in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1463, "options": [ "A. [0.158, 0.112, 0.964, 0.958]", "B. [0.064, 0.166, 0.633, 0.991]", "C. [0.255, 0.435, 0.831, 1.0]", "D. [0.064, 0.425, 0.426, 1.0]" ], "question": "Where is the person currently performing the cut located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1483, "options": [ "A. [0.247, 0.079, 0.619, 0.832]", "B. [0.103, 0.0, 0.466, 0.65]", "C. [0.483, 0.223, 0.676, 0.86]", "D. [0.201, 0.325, 0.367, 0.984]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1612, "options": [ "A. [0.344, 0.256, 0.636, 0.949]", "B. [0.391, 0.423, 0.894, 1.0]", "C. [0.112, 0.229, 0.622, 0.686]", "D. [0.252, 0.021, 1.0, 0.964]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1662, "options": [ "A. [0.556, 0.746, 0.615, 0.933]", "B. [0.644, 0.786, 0.858, 0.712]", "C. [0.203, 0.095, 0.787, 0.921]", "D. [0.291, 0.571, 0.526, 0.904]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1697, "options": [ "A. talk to (e.g., self, a person, a group) [0.152, 0.157, 0.75, 0.936] -> carry/hold (an object) [0.132, 0.133, 0.709, 0.961] -> touch (an object) [0.132, 0.133, 0.709, 0.961]", "B. ride (e.g., a bike, a car, a horse) [0.302, 0.080, 0.924, 1.000] -> carry/hold (an object) [0.174, 0.191, 0.541, 0.767] -> touch (an object) [0.144, 0.234, 0.893, 1.000]", "C. carry/hold (an object) [0.000, 0.119, 0.891, 0.948] -> talk to (e.g., self, a person, a group) [0.291, 0.004, 0.744, 1.000] -> ride (e.g., a bike, a car, a horse) [0.126, 0.000, 0.737, 1.000]", "D. get up [0.270, 0.308, 0.721, 0.913] -> talk to (e.g., self, a person, a group) [0.152, 0.072, 0.892, 1.000] -> touch (an object) [0.093, 0.097, 0.540, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.132, 0.133, 0.709, 0.961] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1699, "options": [ "A. touch (an object) [0.125, 0.139, 0.968, 0.878] -> talk to (e.g., self, a person, a group) [0.286, 0.285, 0.620, 1.000] -> open (e.g., a window, a car door) [0.178, 0.148, 0.538, 1.000] -> play with pets [0.267, 0.364, 0.406, 0.912]", "B. touch (an object) [0.179, 0.162, 0.789, 0.977] -> talk to (e.g., self, a person, a group) [0.179, 0.162, 0.789, 0.977] -> carry/hold (an object) [0.157, 0.215, 0.597, 0.968] -> play with pets [0.157, 0.215, 0.597, 0.968]", "C. play with pets [0.213, 0.334, 0.745, 0.807] -> touch (an object) [0.280, 0.283, 0.689, 1.000] -> shoot [0.347, 0.282, 0.681, 0.862] -> talk to (e.g., self, a person, a group) [0.212, 0.165, 0.798, 1.000]", "D. sing to (e.g., self, a person, a group) [0.012, 0.107, 0.955, 0.825] -> talk to (e.g., self, a person, a group) [0.181, 0.213, 0.798, 1.000] -> carry/hold (an object) [0.104, 0.269, 0.735, 1.000] -> play with pets [0.034, 0.151, 0.481, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.157, 0.215, 0.597, 0.968] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1715, "options": [ "A. talk to (e.g., self, a person, a group) -> walk -> crouch/kneel -> touch (an object) -> play with pets -> sing to (e.g., self, a person, a group) -> carry/hold (an object)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> touch (an object) -> play with pets -> walk -> bend/bow (at the waist) -> crouch/kneel", "C. carry/hold (an object) -> bend/bow (at the waist) -> talk to (e.g., self, a person, a group) -> crouch/kneel -> touch (an object) -> crouch/kneel -> play with pets", "D. dress/put on clothing -> touch (an object) -> walk -> bend/bow (at the waist) -> play with pets -> talk to (e.g., self, a person, a group) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1760, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.245, 0.076, 0.892, 0.968] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1760 ]

[ { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 932, "options": [ "A. [0.022, 0.194, 0.573, 0.998]", "B. [0.171, 0.0, 0.748, 0.73]", "C. [0.603, 0.0, 1.0, 0.975]", "D. [0.377, 0.123, 0.86, 0.987]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 950, "options": [ "A. talk to (e.g., self, a person, a group) [0.482, 0.119, 0.683, 0.849] -> walk [0.423, 0.149, 0.639, 0.967] -> carry/hold (an object) [0.423, 0.149, 0.639, 0.967]", "B. walk [0.445, 0.006, 0.543, 1.000] -> talk to (e.g., self, a person, a group) [0.333, 0.000, 0.672, 0.916] -> play board game [0.263, 0.294, 0.475, 0.770]", "C. put down [0.240, 0.129, 0.589, 0.802] -> talk to (e.g., self, a person, a group) [0.507, 0.249, 0.803, 0.654] -> walk [0.405, 0.017, 0.470, 1.000]", "D. talk to (e.g., self, a person, a group) [0.541, 0.199, 0.559, 0.837] -> carry/hold (an object) [0.449, 0.128, 0.719, 0.814] -> take a photo [0.575, 0.000, 0.784, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.423, 0.149, 0.639, 0.967] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 951, "options": [ "A. sail boat [0.602, 0.067, 0.812, 0.959] -> walk [0.385, 0.271, 0.635, 0.903] -> talk to (e.g., self, a person, a group) [0.363, 0.000, 0.624, 1.000]", "B. work on a computer [0.310, 0.045, 0.722, 0.784] -> talk to (e.g., self, a person, a group) [0.588, 0.000, 0.873, 0.951] -> carry/hold (an object) [0.517, 0.096, 0.449, 0.776]", "C. walk [0.232, 0.227, 0.592, 1.000] -> turn (e.g., a screwdriver) [0.626, 0.143, 0.505, 0.674] -> carry/hold (an object) [0.494, 0.000, 0.794, 0.930]", "D. talk to (e.g., self, a person, a group) [0.482, 0.119, 0.683, 0.849] -> walk [0.423, 0.149, 0.639, 0.967] -> carry/hold (an object) [0.423, 0.149, 0.639, 0.967]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.423, 0.149, 0.639, 0.967] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 957, "options": [ "A. [0.745, 0.391, 0.568, 1.0]", "B. [0.633, 0.139, 0.906, 0.905]", "C. [0.046, 0.126, 0.479, 0.994]", "D. [0.482, 0.119, 0.683, 0.849]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 966, "options": [ "A. talk to (e.g., self, a person, a group) [0.723, 0.05, 0.888, 0.856] -> walk [0.755, 0.052, 0.892, 0.731] -> carry/hold (an object) [0.755, 0.052, 0.892, 0.731] -> hand wave [0.755, 0.052, 0.892, 0.731]", "B. talk to (e.g., self, a person, a group) [0.744, 0.186, 0.913, 0.859] -> hug (a person) [0.668, 0.000, 0.853, 0.806] -> hand wave [0.630, 0.085, 1.000, 0.589] -> walk [0.889, 0.044, 0.962, 0.812]", "C. talk to (e.g., self, a person, a group) [0.721, 0.000, 0.783, 0.713] -> carry/hold (an object) [0.803, 0.000, 0.927, 0.683] -> hand wave [0.765, 0.209, 1.000, 0.683] -> grab (a person) [0.658, 0.076, 0.717, 0.569]", "D. run/jog [0.833, 0.017, 0.881, 0.739] -> talk to (e.g., self, a person, a group) [0.588, 0.042, 0.826, 0.731] -> hand wave [0.877, 0.153, 1.000, 0.615] -> walk [0.568, 0.145, 0.905, 0.722]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.755, 0.052, 0.892, 0.731] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1024, "options": [ "A. read [0.000, 0.152, 0.133, 1.000] -> talk to (e.g., self, a person, a group) [0.202, 0.151, 0.217, 1.000] -> touch (an object) [0.016, 0.341, 0.383, 0.857]", "B. enter [0.000, 0.201, 0.209, 0.923] -> eat [0.061, 0.244, 0.475, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.070, 0.125, 1.000]", "C. talk to (e.g., self, a person, a group) [0.036, 0.293, 0.307, 1.000] -> touch (an object) [0.033, 0.121, 0.473, 0.847] -> chop [0.000, 0.369, 0.481, 0.882]", "D. talk to (e.g., self, a person, a group) [0.002, 0.248, 0.317, 0.989] -> eat [0.002, 0.251, 0.325, 0.996] -> touch (an object) [0.002, 0.251, 0.325, 0.996]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.002, 0.251, 0.325, 0.996] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1028, "options": [ "A. walk [0.406, 0.135, 0.474, 0.997] -> fall down [0.500, 0.000, 0.762, 1.000] -> talk to (e.g., self, a person, a group) [0.385, 0.142, 0.734, 0.908]", "B. talk to (e.g., self, a person, a group) [0.511, 0.025, 0.596, 0.781] -> stir [0.374, 0.025, 0.436, 0.862] -> carry/hold (an object) [0.317, 0.075, 0.600, 0.987]", "C. walk [0.377, 0.101, 0.576, 0.97] -> carry/hold (an object) [0.377, 0.101, 0.576, 0.97] -> talk to (e.g., self, a person, a group) [0.377, 0.101, 0.576, 0.97]", "D. lie/sleep [0.220, 0.049, 0.515, 1.000] -> walk [0.569, 0.000, 0.567, 1.000] -> carry/hold (an object) [0.190, 0.000, 0.644, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.377, 0.101, 0.576, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1081, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.809, 0.214, 0.93, 0.57] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1150, "options": [ "A. sit, touch (an object)", "B. sit, touch (an object), watch (a person)", "C. sit, listen to (a person), watch (a person)", "D. sit, touch (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.019, 0.032, 0.594, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1207, "options": [ "A. [0.226, 0.046, 0.767, 0.98]", "B. [0.015, 0.224, 0.474, 0.864]", "C. [0.577, 0.098, 0.705, 0.532]", "D. [0.494, 0.24, 0.823, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1255, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1352, "options": [ "A. 3", "B. 4", "C. 5", "D. 7" ], "question": "How many people were performing the touch (an object) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1470, "options": [ "A. [0.795, 0.18, 0.829, 0.763]", "B. [0.312, 0.205, 0.54, 0.982]", "C. [0.505, 0.241, 0.575, 0.557]", "D. [0.746, 0.464, 0.681, 0.765]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1515, "options": [ "A. [0.012, 0.152, 0.556, 1.0]", "B. [0.169, 0.408, 0.709, 0.729]", "C. [0.368, 0.152, 0.934, 0.995]", "D. [0.629, 0.0, 1.0, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1572, "options": [ "A. [0.019, 0.144, 0.535, 1.0]", "B. [0.0, 0.375, 0.774, 1.0]", "C. [0.0, 0.001, 0.824, 0.979]", "D. [0.0, 0.0, 0.638, 0.736]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1646, "options": [ "A. touch (an object)", "B. text on/look at a cellphone", "C. eat", "D. take a photo" ], "question": "What action is the person currently in the [0.412, 0.206, 0.665, 0.63] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1648, "options": [ "A. sit, touch (an object), listen to (a person)", "B. sit, watch (a person)", "C. stand, carry/hold (an object)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.083, 0.256, 0.295, 0.921] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1685, "options": [ "A. 5", "B. 3", "C. 4", "D. 6" ], "question": "How many people were performing the sit in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1793, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1793 ]

[ { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 946, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)", "C. close (e.g., a door, a box) -> carry/hold (an object)", "D. talk to (e.g., self, a person, a group) -> play with kids" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 971, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. enter -> carry/hold (an object)", "C. get up -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 993, "options": [ "A. [0.212, 0.005, 0.858, 0.978]", "B. [0.335, 0.039, 1.0, 0.778]", "C. [0.572, 0.417, 0.823, 0.997]", "D. [0.136, 0.299, 0.641, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 996, "options": [ "A. [0.745, 0.139, 0.876, 0.973]", "B. [0.086, 0.259, 0.384, 0.946]", "C. [0.344, 0.527, 0.428, 0.889]", "D. [0.16, 0.792, 0.141, 0.699]" ], "question": "Where was the person currently performing the crouch/kneel in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 996, "options": [ "A. bend/bow (at the waist) [0.527, 0.357, 0.662, 0.987] -> carry/hold (an object) [0.527, 0.357, 0.662, 0.987] -> talk to (e.g., self, a person, a group) [0.527, 0.357, 0.662, 0.987]", "B. bend/bow (at the waist) [0.719, 0.384, 0.676, 0.976] -> dance [0.548, 0.354, 0.556, 0.866] -> talk to (e.g., self, a person, a group) [0.534, 0.357, 0.763, 1.000]", "C. clink glass [0.664, 0.521, 0.678, 1.000] -> talk to (e.g., self, a person, a group) [0.578, 0.312, 0.592, 1.000] -> carry/hold (an object) [0.647, 0.168, 0.472, 0.903]", "D. bend/bow (at the waist) [0.622, 0.382, 0.512, 0.861] -> carry/hold (an object) [0.332, 0.349, 0.541, 0.848] -> kick (an object) [0.375, 0.416, 0.620, 0.830]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.527, 0.357, 0.662, 0.987] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1006, "options": [ "A. talk to (e.g., self, a person, a group) [0.032, 0.114, 0.458, 1.000] -> ride (e.g., a bike, a car, a horse) [0.174, 0.163, 0.467, 0.997] -> carry/hold (an object) [0.231, 0.492, 0.168, 1.000]", "B. talk to (e.g., self, a person, a group) [0.000, 0.417, 0.437, 0.943] -> walk [0.147, 0.194, 0.463, 1.000] -> put down [0.174, 0.262, 0.200, 0.870]", "C. walk [0.122, 0.307, 0.339, 0.997] -> carry/hold (an object) [0.122, 0.307, 0.339, 0.997] -> talk to (e.g., self, a person, a group) [0.122, 0.307, 0.339, 0.997]", "D. carry/hold (an object) [0.000, 0.420, 0.369, 1.000] -> swim [0.170, 0.494, 0.479, 1.000] -> walk [0.192, 0.139, 0.147, 0.894]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.122, 0.307, 0.339, 0.997] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1029, "options": [ "A. sit, carry/hold (an object), listen to (a person)", "B. stand, watch (a person)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.122, 0.003, 0.413, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1181, "options": [ "A. [0.233, 0.107, 0.619, 1.0]", "B. [0.125, 0.0, 0.436, 1.0]", "C. [0.344, 0.527, 0.428, 0.889]", "D. [0.395, 0.329, 0.721, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1190, "options": [ "A. [0.552, 0.003, 0.878, 0.99]", "B. [0.254, 0.0, 0.86, 0.994]", "C. [0.765, 0.196, 1.0, 1.0]", "D. [0.397, 0.291, 0.888, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1202, "options": [ "A. carry/hold (an object), crouch/kneel", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, carry/hold (an object), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.54, 0.451, 0.69, 0.991] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1277, "options": [ "A. [0.469, 0.0, 0.624, 0.052]", "B. [0.16, 0.197, 0.245, 0.487]", "C. [0.653, 0.103, 0.687, 0.305]", "D. [0.522, 0.0, 0.466, 0.378]" ], "question": "What location in the frame is the person currently in the [0.128, 0.229, 0.228, 0.823] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1287, "options": [ "A. [0.896, 0.31, 0.667, 0.387]", "B. [0.389, 0.0, 0.569, 0.288]", "C. [0.631, 0.104, 0.681, 0.35]", "D. [0.383, 0.304, 0.642, 0.991]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1351, "options": [ "A. stand, listen to (a person), watch (a person)", "B. walk, watch (a person)", "C. stand, listen to (a person)", "D. stand, catch (an object)" ], "question": "What action is the person at the location [0.307, 0.015, 0.48, 0.982] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1357, "options": [ "A. bend/bow (at the waist) [0.389, 0.269, 0.511, 0.851] -> play with pets [0.495, 0.378, 0.596, 1.000] -> carry/hold (an object) [0.258, 0.230, 0.454, 1.000]", "B. carry/hold (an object) [0.432, 0.115, 0.823, 1.000] -> bend/bow (at the waist) [0.532, 0.345, 0.818, 0.899] -> lift (a person) [0.300, 0.409, 0.777, 1.000]", "C. bend/bow (at the waist) [0.380, 0.163, 0.623, 0.866] -> carry/hold (an object) [0.430, 0.475, 0.478, 1.000] -> point to (an object) [0.183, 0.482, 0.561, 1.000]", "D. bend/bow (at the waist) [0.383, 0.304, 0.642, 0.991] -> carry/hold (an object) [0.383, 0.304, 0.642, 0.991] -> talk to (e.g., self, a person, a group) [0.383, 0.304, 0.642, 0.991]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.383, 0.304, 0.642, 0.991] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1361, "options": [ "A. row boat", "B. bend/bow (at the waist) and talk to (e.g., self, a person, a group)", "C. clink glass", "D. carry/hold (an object) and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.488, 0.007, 0.878, 0.994] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1363, "options": [ "A. talk to (e.g., self, a person, a group) [0.548, 0.141, 0.714, 0.989] -> drink [0.490, 0.086, 0.883, 0.928] -> bend/bow (at the waist) [0.623, 0.135, 0.727, 1.000]", "B. read [0.302, 0.025, 0.925, 1.000] -> carry/hold (an object) [0.312, 0.192, 0.938, 1.000] -> talk to (e.g., self, a person, a group) [0.343, 0.039, 0.974, 0.819]", "C. talk to (e.g., self, a person, a group) [0.619, 0.000, 0.807, 1.000] -> hand shake [0.602, 0.043, 1.000, 1.000] -> bend/bow (at the waist) [0.424, 0.000, 1.000, 1.000]", "D. bend/bow (at the waist) [0.46, 0.013, 0.878, 0.998] -> carry/hold (an object) [0.46, 0.013, 0.878, 0.998] -> talk to (e.g., self, a person, a group) [0.46, 0.013, 0.878, 0.998]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.46, 0.013, 0.878, 0.998] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1368, "options": [ "A. talk to (e.g., self, a person, a group) [0.311, 0.000, 1.000, 0.803] -> bend/bow (at the waist) [0.177, 0.000, 0.987, 1.000] -> shoot [0.148, 0.000, 0.974, 0.909]", "B. read [0.108, 0.116, 0.828, 1.000] -> carry/hold (an object) [0.421, 0.184, 0.768, 1.000] -> talk to (e.g., self, a person, a group) [0.407, 0.000, 0.893, 0.829]", "C. talk to (e.g., self, a person, a group) [0.242, 0.000, 0.959, 1.000] -> bend/bow (at the waist) [0.157, 0.156, 0.978, 1.000] -> paint [0.198, 0.065, 1.000, 0.983]", "D. bend/bow (at the waist) [0.263, 0.0, 0.847, 1.0] -> carry/hold (an object) [0.263, 0.0, 0.847, 1.0] -> talk to (e.g., self, a person, a group) [0.263, 0.0, 0.847, 1.0]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.263, 0.0, 0.847, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1419, "options": [ "A. talk to (e.g., self, a person, a group) -> eat", "B. stir -> talk to (e.g., self, a person, a group)", "C. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "D. bend/bow (at the waist) -> catch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1506, "options": [ "A. [0.27, 0.019, 0.875, 1.0]", "B. [0.508, 0.045, 1.0, 1.0]", "C. [0.307, 0.338, 0.539, 0.996]", "D. [0.009, 0.0, 0.747, 0.985]" ], "question": "Where is the person currently performing the hug (a person) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1507, "options": [ "A. exit -> talk to (e.g., self, a person, a group)", "B. lift/pick up -> hug (a person)", "C. hug (a person) -> talk to (e.g., self, a person, a group)", "D. turn (e.g., a screwdriver) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1610, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1666, "options": [ "A. stand, listen to (a person), watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. stand, watch (a person)", "D. stand, give/serve (an object) to (a person), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.122, 0.434, 0.313, 0.997] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1773, "options": [ "A. sing to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)", "B. lie/sleep -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> fight/hit (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1773 ]

[ { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 921, "options": [ "A. jump/leap and talk to (e.g., self, a person, a group)", "B. text on/look at a cellphone", "C. answer phone", "D. paint" ], "question": "What action is the person currently in the [0.13, 0.151, 0.407, 0.743] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 948, "options": [ "A. [0.065, 0.439, 0.414, 0.893]", "B. [0.056, 0.0, 1.0, 0.676]", "C. [0.0, 0.122, 0.796, 1.0]", "D. [0.199, 0.138, 0.933, 0.891]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 981, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.296, 0.219, 0.59, 0.905] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1060, "options": [ "A. [0.286, 0.0, 0.917, 0.626]", "B. [0.153, 0.21, 0.881, 0.88]", "C. [0.584, 0.129, 0.939, 0.885]", "D. [0.824, 0.18, 1.0, 1.0]" ], "question": "Where was the person currently performing the answer phone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1068, "options": [ "A. play board game -> answer phone -> text on/look at a cellphone", "B. answer phone -> paint -> text on/look at a cellphone", "C. carry/hold (an object) -> text on/look at a cellphone -> answer phone", "D. swim -> text on/look at a cellphone -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1090, "options": [ "A. [0.495, 0.24, 0.854, 0.875]", "B. [0.0, 0.377, 0.809, 1.0]", "C. [0.16, 0.162, 0.639, 0.888]", "D. [0.175, 0.0, 0.536, 0.609]" ], "question": "Where is the person currently performing the answer phone located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1133, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1197, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1208, "options": [ "A. sing to (e.g., self, a person, a group) -> walk", "B. talk to (e.g., self, a person, a group) -> walk", "C. lift (a person) -> walk", "D. carry/hold (an object) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1276, "options": [ "A. [0.499, 0.282, 0.964, 0.877]", "B. [0.0, 0.39, 0.765, 0.787]", "C. [0.027, 0.001, 1.0, 0.569]", "D. [0.259, 0.14, 0.819, 0.862]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1335, "options": [ "A. turn (e.g., a screwdriver) -> eat -> carry/hold (an object) -> walk", "B. carry/hold (an object) -> cut -> eat -> walk", "C. walk -> carry/hold (an object) -> eat -> climb (e.g., a mountain)", "D. eat -> talk to (e.g., self, a person, a group) -> cut -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1339, "options": [ "A. fight/hit (a person) -> cut", "B. sing to (e.g., self, a person, a group) -> carry/hold (an object)", "C. carry/hold (an object) -> cut", "D. lift (a person) -> cut" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1348, "options": [ "A. [0.166, 0.151, 0.655, 0.883]", "B. [0.278, 0.337, 0.561, 1.0]", "C. [0.269, 0.456, 0.885, 1.0]", "D. [0.484, 0.164, 0.797, 0.883]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1352, "options": [ "A. [0.647, 0.0, 1.0, 0.751]", "B. [0.069, 0.241, 0.465, 0.874]", "C. [0.089, 0.384, 0.996, 1.0]", "D. [0.375, 0.139, 0.923, 0.87]" ], "question": "Where is the person currently performing the bend/bow (at the waist) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1400, "options": [ "A. [0.023, 0.121, 0.64, 0.872]", "B. [0.34, 0.147, 0.965, 0.868]", "C. [0.174, 0.211, 0.791, 0.634]", "D. [0.518, 0.341, 0.908, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1483, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1518, "options": [ "A. [0.424, 0.445, 0.559, 0.869]", "B. [0.44, 0.0, 1.0, 0.844]", "C. [0.415, 0.239, 0.871, 0.867]", "D. [0.134, 0.123, 0.711, 0.572]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1736, "options": [ "A. [0.222, 0.222, 0.957, 0.883]", "B. [0.178, 0.197, 0.605, 0.862]", "C. [0.492, 0.482, 0.953, 0.92]", "D. [0.0, 0.091, 0.962, 0.7]" ], "question": "Where was the person currently performing the lie/sleep in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1749, "options": [ "A. carry/hold (an object) -> lie/sleep", "B. take (an object) from (a person) -> carry/hold (an object)", "C. eat -> carry/hold (an object)", "D. lie/sleep -> exit" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1790, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, hand clap, watch (a person)", "C. sit, listen to (a person)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.055, 0.185, 0.654, 0.89] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 900, 1790 ]

[ { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1011, "options": [ "A. dance, play musical instrument, listen to (a person)", "B. stand, talk to (e.g., self, a person, a group)", "C. stand, watch (a person)", "D. sit, watch (a person)" ], "question": "What action is the person at the location [0.276, 0.097, 0.426, 0.576] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1048, "options": [ "A. [0.283, 0.163, 0.615, 0.755]", "B. [0.514, 0.227, 0.845, 0.517]", "C. [0.567, 0.086, 0.462, 0.484]", "D. [0.655, 0.231, 1.0, 0.746]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1060, "options": [ "A. talk to (e.g., self, a person, a group) [0.613, 0.231, 0.998, 0.734] -> carry/hold (an object) [0.581, 0.228, 0.965, 0.756] -> give/serve (an object) to (a person) [0.581, 0.228, 0.965, 0.756]", "B. carry/hold (an object) [0.602, 0.282, 1.000, 0.621] -> talk to (e.g., self, a person, a group) [0.684, 0.060, 0.943, 0.642] -> talk to (e.g., self, a person, a group) [0.484, 0.041, 0.871, 0.689]", "C. give/serve (an object) to (a person) [0.759, 0.132, 0.886, 0.742] -> row boat [0.491, 0.343, 0.847, 0.789] -> talk to (e.g., self, a person, a group) [0.687, 0.218, 0.968, 0.611]", "D. talk to (e.g., self, a person, a group) [0.741, 0.035, 0.903, 0.810] -> hand wave [0.727, 0.104, 0.790, 0.669] -> give/serve (an object) to (a person) [0.718, 0.167, 0.842, 0.928]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.581, 0.228, 0.965, 0.756] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1075, "options": [ "A. push (another person) [0.304, 0.099, 0.685, 0.851] -> sing to (e.g., self, a person, a group) [0.000, 0.044, 0.911, 1.000] -> carry/hold (an object) [0.127, 0.000, 0.807, 1.000]", "B. sing to (e.g., self, a person, a group) [0.087, 0.154, 0.803, 0.964] -> carry/hold (an object) [0.158, 0.12, 0.761, 0.982] -> dance [0.158, 0.12, 0.761, 0.982]", "C. dance [0.000, 0.000, 0.721, 0.883] -> carry/hold (an object) [0.297, 0.000, 0.940, 1.000] -> hand wave [0.147, 0.167, 0.641, 1.000]", "D. carry/hold (an object) [0.063, 0.035, 0.766, 1.000] -> close (e.g., a door, a box) [0.288, 0.086, 0.675, 0.970] -> sing to (e.g., self, a person, a group) [0.264, 0.203, 0.813, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.158, 0.12, 0.761, 0.982] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1107, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1115, "options": [ "A. play musical instrument and sing to (e.g., self, a person, a group)", "B. walk", "C. bend/bow (at the waist) and carry/hold (an object)", "D. carry/hold (an object)" ], "question": "What action is the person currently in the [0.091, 0.254, 0.328, 0.611] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1117, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, watch (a person)", "C. stand, play musical instrument, listen to (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.094, 0.239, 0.328, 0.586] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1120, "options": [ "A. talk to (e.g., self, a person, a group) [0.741, 0.177, 1.000, 0.746] -> bend/bow (at the waist) [0.830, 0.290, 0.855, 0.553] -> clink glass [0.516, 0.308, 0.893, 0.867]", "B. jump/leap [0.527, 0.147, 1.000, 0.599] -> bend/bow (at the waist) [0.802, 0.294, 0.930, 0.790] -> talk to (e.g., self, a person, a group) [0.791, 0.117, 1.000, 0.838]", "C. drive (e.g., a car, a truck) [0.519, 0.359, 0.892, 0.564] -> bend/bow (at the waist) [0.858, 0.277, 0.992, 0.676] -> talk to (e.g., self, a person, a group) [0.520, 0.166, 0.867, 0.831]", "D. bend/bow (at the waist) [0.715, 0.245, 0.983, 0.748] -> carry/hold (an object) [0.715, 0.245, 0.983, 0.748] -> talk to (e.g., self, a person, a group) [0.715, 0.245, 0.983, 0.748]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.715, 0.245, 0.983, 0.748] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1158, "options": [ "A. cook [0.203, 0.456, 0.630, 0.961] -> carry/hold (an object) [0.377, 0.243, 0.636, 0.767] -> stir [0.314, 0.350, 0.441, 0.905]", "B. bend/bow (at the waist) [0.23, 0.307, 0.576, 0.815] -> carry/hold (an object) [0.23, 0.307, 0.576, 0.815] -> cook [0.23, 0.307, 0.576, 0.815]", "C. bend/bow (at the waist) [0.048, 0.413, 0.773, 0.856] -> carry/hold (an object) [0.142, 0.264, 0.758, 1.000] -> open (e.g., a window, a car door) [0.223, 0.357, 0.381, 0.923]", "D. cook [0.073, 0.267, 0.430, 0.932] -> carry/hold (an object) [0.208, 0.168, 0.516, 0.942] -> play musical instrument [0.245, 0.367, 0.393, 0.932]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.23, 0.307, 0.576, 0.815] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1214, "options": [ "A. [0.721, 0.445, 0.929, 0.965]", "B. [0.555, 0.0, 0.495, 0.756]", "C. [0.13, 0.0, 0.492, 0.852]", "D. [0.283, 0.107, 0.529, 0.996]" ], "question": "Where was the person currently performing the sing to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1318, "options": [ "A. [0.074, 0.142, 0.503, 0.934]", "B. [0.259, 0.132, 1.0, 1.0]", "C. [0.311, 0.0, 0.838, 0.769]", "D. [0.047, 0.034, 0.872, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1384, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1409, "options": [ "A. [0.074, 0.14, 0.57, 0.961]", "B. [0.735, 0.513, 0.948, 0.878]", "C. [0.363, 0.0, 0.618, 1.0]", "D. [0.034, 0.058, 0.406, 0.789]" ], "question": "Where was the person currently performing the play musical instrument in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1468, "options": [ "A. [0.0, 0.099, 0.888, 0.991]", "B. [0.0, 0.32, 0.621, 1.0]", "C. [0.406, 0.044, 0.951, 0.929]", "D. [0.068, 0.0, 0.974, 0.82]" ], "question": "Where is the person currently performing the sing to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1494, "options": [ "A. [0.444, 0.302, 0.837, 0.942]", "B. [0.271, 0.072, 0.679, 0.766]", "C. [0.064, 0.076, 0.564, 0.981]", "D. [0.05, 0.037, 0.949, 0.474]" ], "question": "Where was the person currently performing the play musical instrument in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1547, "options": [ "A. [0.171, 0.323, 0.253, 0.625]", "B. [0.402, 0.501, 0.423, 0.738]", "C. [0.233, 0.13, 0.889, 0.97]", "D. [0.077, 0.445, 0.225, 0.807]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1677, "options": [ "A. watch (e.g., TV) [0.333, 0.167, 1.000, 0.800] -> talk to (e.g., self, a person, a group) [0.275, 0.026, 0.657, 0.745] -> carry/hold (an object) [0.386, 0.294, 0.662, 0.929]", "B. give/serve (an object) to (a person) [0.420, 0.280, 0.843, 0.803] -> martial art [0.588, 0.024, 0.805, 0.711] -> carry/hold (an object) [0.523, 0.183, 0.756, 0.841]", "C. carry/hold (an object) [0.527, 0.236, 0.763, 0.990] -> talk to (e.g., self, a person, a group) [0.594, 0.110, 0.913, 0.850] -> sail boat [0.277, 0.309, 0.870, 1.000]", "D. carry/hold (an object) [0.412, 0.152, 0.823, 0.888] -> give/serve (an object) to (a person) [0.412, 0.152, 0.823, 0.888] -> talk to (e.g., self, a person, a group) [0.412, 0.152, 0.823, 0.888]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.412, 0.152, 0.823, 0.888] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1707, "options": [ "A. [0.199, 0.0, 1.0, 0.779]", "B. [0.277, 0.157, 0.542, 0.568]", "C. [0.064, 0.25, 0.68, 0.93]", "D. [0.104, 0.108, 0.967, 0.99]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1798, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1798 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 956, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 960, "options": [ "A. bend/bow (at the waist) -> smoke", "B. talk to (e.g., self, a person, a group) -> bend/bow (at the waist)", "C. talk to (e.g., self, a person, a group) -> pull (an object)", "D. talk to (e.g., self, a person, a group) -> kick (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 962, "options": [ "A. [0.193, 0.0, 0.797, 1.0]", "B. [0.455, 0.277, 1.0, 1.0]", "C. [0.174, 0.0, 1.0, 0.718]", "D. [0.33, 0.0, 0.992, 0.993]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1008, "options": [ "A. walk -> carry/hold (an object)", "B. carry/hold (an object) -> dig", "C. walk -> lift/pick up", "D. dress/put on clothing -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1031, "options": [ "A. talk to (e.g., self, a person, a group) -> throw", "B. play board game -> carry/hold (an object)", "C. carry/hold (an object) -> stir", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1062, "options": [ "A. [0.772, 0.344, 0.416, 0.758]", "B. [0.582, 0.0, 0.791, 0.761]", "C. [0.477, 0.057, 0.569, 0.464]", "D. [0.025, 0.012, 0.873, 0.984]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1084, "options": [ "A. walk and carry/hold (an object) and talk to (e.g., self, a person, a group)", "B. walk and carry/hold (an object)", "C. shoot", "D. hand wave" ], "question": "What action is the person currently in the [0.513, 0.026, 0.627, 0.411] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1129, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1207, "options": [ "A. [0.083, 0.003, 0.95, 0.991]", "B. [0.0, 0.274, 0.675, 0.794]", "C. [0.193, 0.0, 0.652, 0.72]", "D. [0.373, 0.012, 0.984, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1284, "options": [ "A. [0.032, 0.018, 0.699, 1.0]", "B. [0.464, 0.0, 0.664, 0.935]", "C. [0.305, 0.009, 0.584, 0.692]", "D. [0.177, 0.056, 0.385, 0.643]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1341, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1347, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1382, "options": [ "A. [0.483, 0.0, 1.0, 0.859]", "B. [0.015, 0.0, 0.545, 0.977]", "C. [0.46, 0.324, 0.827, 1.0]", "D. [0.2, 0.027, 0.954, 0.974]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1501, "options": [ "A. [0.015, 0.003, 0.517, 0.983]", "B. [0.344, 0.0, 1.0, 0.954]", "C. [0.364, 0.277, 0.883, 1.0]", "D. [0.193, 0.0, 0.797, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1512, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1527, "options": [ "A. [0.056, 0.002, 0.686, 0.993]", "B. [0.0, 0.298, 0.482, 0.953]", "C. [0.0, 0.0, 0.444, 0.847]", "D. [0.224, 0.045, 0.829, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1596, "options": [ "A. [0.072, 0.0, 0.916, 0.705]", "B. [0.0, 0.0, 0.481, 0.718]", "C. [0.473, 0.0, 0.704, 0.951]", "D. [0.057, 0.012, 0.74, 1.0]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1627, "options": [ "A. [0.0, 0.283, 0.379, 1.0]", "B. [0.32, 0.336, 0.752, 1.0]", "C. [0.47, 0.0, 0.67, 0.955]", "D. [0.112, 0.043, 0.674, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1676, "options": [ "A. walk, listen to (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, listen to (a person), watch (a person)", "D. carry/hold (an object), get up, listen to (a person)" ], "question": "What action is the person at the location [0.186, 0.01, 0.496, 0.995] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1712, "options": [ "A. open (e.g., a window, a car door) -> walk", "B. walk -> talk to (e.g., self, a person, a group)", "C. walk -> play board game", "D. talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1793, "options": [ "A. walk -> carry/hold (an object)", "B. play musical instrument -> walk", "C. carry/hold (an object) -> get up", "D. ride (e.g., a bike, a car, a horse) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1793 ]

[ { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 925, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.132, 0.155, 0.863, 0.965] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 993, "options": [ "A. dance -> lie/sleep", "B. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> lie/sleep", "D. talk to (e.g., self, a person, a group) -> run/jog" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1046, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, watch (a person)", "C. stand, touch (an object), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.131, 0.301, 0.526, 0.989] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1062, "options": [ "A. [0.0, 0.468, 0.43, 0.885]", "B. [0.605, 0.219, 0.773, 0.975]", "C. [0.127, 0.19, 0.668, 0.961]", "D. [0.232, 0.014, 0.613, 0.775]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1103, "options": [ "A. 6", "B. 4", "C. 3", "D. 5" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1140, "options": [ "A. stand, carry/hold (an object), watch (a person)", "B. walk, listen to (a person), watch (a person)", "C. stand, watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.121, 0.007, 0.543, 0.975] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1174, "options": [ "A. run/jog [0.159, 0.176, 0.472, 0.860] -> talk to (e.g., self, a person, a group) [0.054, 0.163, 0.677, 0.949] -> carry/hold (an object) [0.021, 0.180, 0.554, 0.880]", "B. talk to (e.g., self, a person, a group) [0.021, 0.000, 0.504, 1.000] -> carry/hold (an object) [0.191, 0.303, 0.412, 0.822] -> grab (a person) [0.124, 0.000, 0.498, 1.000]", "C. chop [0.301, 0.022, 0.400, 0.798] -> carry/hold (an object) [0.014, 0.220, 0.412, 0.898] -> talk to (e.g., self, a person, a group) [0.094, 0.029, 0.415, 0.905]", "D. carry/hold (an object) [0.168, 0.105, 0.493, 0.97] -> smoke [0.168, 0.105, 0.493, 0.97] -> talk to (e.g., self, a person, a group) [0.168, 0.105, 0.493, 0.97]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.168, 0.105, 0.493, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1176, "options": [ "A. carry/hold (an object) [0.168, 0.105, 0.493, 0.97] -> smoke [0.168, 0.105, 0.493, 0.97] -> talk to (e.g., self, a person, a group) [0.168, 0.105, 0.493, 0.97]", "B. hand clap [0.028, 0.182, 0.504, 0.898] -> smoke [0.117, 0.219, 0.552, 0.872] -> talk to (e.g., self, a person, a group) [0.181, 0.204, 0.382, 0.984]", "C. smoke [0.296, 0.164, 0.515, 1.000] -> talk to (e.g., self, a person, a group) [0.137, 0.000, 0.488, 1.000] -> hug (a person) [0.215, 0.084, 0.472, 0.973]", "D. kick (an object) [0.212, 0.247, 0.583, 1.000] -> smoke [0.184, 0.074, 0.562, 1.000] -> talk to (e.g., self, a person, a group) [0.092, 0.121, 0.383, 0.901]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.168, 0.105, 0.493, 0.97] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1238, "options": [ "A. lie/sleep -> walk", "B. talk to (e.g., self, a person, a group) -> lie/sleep", "C. talk to (e.g., self, a person, a group) -> smoke", "D. touch (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1266, "options": [ "A. clink glass -> carry/hold (an object)", "B. row boat -> carry/hold (an object)", "C. carry/hold (an object) -> smoke", "D. smoke -> play with pets" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1269, "options": [ "A. lie/sleep -> crouch/kneel", "B. talk to (e.g., self, a person, a group) -> lie/sleep", "C. talk to (e.g., self, a person, a group) -> take (an object) from (a person)", "D. lie/sleep -> catch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1273, "options": [ "A. 3", "B. 6", "C. 4", "D. 5" ], "question": "How many people were performing the stand in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1277, "options": [ "A. [0.382, 0.583, 0.519, 0.837]", "B. [0.468, 0.03, 1.0, 1.0]", "C. [0.235, 0.0, 0.854, 0.97]", "D. [0.07, 0.0, 0.675, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1354, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1409, "options": [ "A. carry/hold (an object) and close (e.g., a door, a box)", "B. lie/sleep", "C. cook", "D. push (an object)" ], "question": "What action is the person currently in the [0.328, 0.211, 0.436, 0.809] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1418, "options": [ "A. [0.433, 0.327, 0.696, 0.975]", "B. [0.16, 0.0, 0.51, 0.594]", "C. [0.069, 0.37, 0.474, 1.0]", "D. [0.32, 0.201, 0.426, 0.813]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1420, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1589, "options": [ "A. play board game [0.343, 0.33, 0.404, 0.698] -> walk [0.316, 0.304, 0.398, 0.704] -> carry/hold (an object) [0.316, 0.304, 0.398, 0.704]", "B. walk [0.312, 0.229, 0.452, 0.745] -> carry/hold (an object) [0.369, 0.363, 0.223, 0.789] -> throw [0.173, 0.196, 0.300, 0.745]", "C. hand shake [0.393, 0.420, 0.389, 0.518] -> walk [0.402, 0.276, 0.551, 0.707] -> carry/hold (an object) [0.148, 0.394, 0.218, 0.698]", "D. play board game [0.357, 0.215, 0.476, 0.672] -> jump/leap [0.481, 0.330, 0.247, 0.631] -> walk [0.273, 0.287, 0.350, 0.649]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.316, 0.304, 0.398, 0.704] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1606, "options": [ "A. carry/hold (an object) [0.459, 0.211, 0.623, 0.996] -> crouch/kneel [0.293, 0.037, 0.701, 1.000] -> bend/bow (at the waist) [0.288, 0.188, 0.977, 0.726]", "B. bend/bow (at the waist) [0.374, 0.233, 0.82, 0.902] -> carry/hold (an object) [0.374, 0.233, 0.82, 0.902] -> play board game [0.374, 0.233, 0.82, 0.902]", "C. play board game [0.266, 0.037, 0.838, 0.844] -> carry/hold (an object) [0.399, 0.089, 0.686, 1.000] -> bend/bow (at the waist) [0.238, 0.296, 0.669, 0.949]", "D. play board game [0.377, 0.196, 0.788, 0.961] -> bend/bow (at the waist) [0.537, 0.102, 0.753, 0.934] -> fight/hit (a person) [0.547, 0.231, 0.849, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.374, 0.233, 0.82, 0.902] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1631, "options": [ "A. stand, carry/hold (an object)", "B. sit, carry/hold (an object), talk to (e.g., self, a person, a group)", "C. stand, watch (a person)", "D. stand, touch (an object)" ], "question": "What action is the person at the location [0.208, 0.28, 0.361, 0.964] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1652, "options": [ "A. ride (e.g., a bike, a car, a horse) -> walk", "B. walk -> carry/hold (an object)", "C. dance -> walk", "D. carry/hold (an object) -> shovel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1707, "options": [ "A. [0.236, 0.098, 0.616, 0.836]", "B. [0.309, 0.128, 0.869, 0.936]", "C. [0.066, 0.0, 0.768, 0.604]", "D. [0.0, 0.157, 0.338, 0.726]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1759, "options": [ "A. [0.291, 0.0, 0.658, 0.735]", "B. [0.377, 0.113, 0.687, 0.934]", "C. [0.308, 0.234, 0.389, 0.783]", "D. [0.619, 0.181, 0.912, 0.883]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1772, "options": [ "A. talk to (e.g., self, a person, a group) [0.202, 0.253, 0.285, 0.676] -> push (another person) [0.180, 0.349, 0.257, 0.644] -> touch (an object) [0.096, 0.273, 0.528, 0.668]", "B. talk to (e.g., self, a person, a group) [0.229, 0.289, 0.597, 1.000] -> carry/hold (an object) [0.260, 0.339, 0.350, 0.927] -> drive (e.g., a car, a truck) [0.037, 0.036, 0.637, 0.744]", "C. carry/hold (an object) [0.223, 0.16, 0.452, 0.803] -> touch (an object) [0.223, 0.16, 0.452, 0.803] -> talk to (e.g., self, a person, a group) [0.223, 0.16, 0.452, 0.803]", "D. carry/hold (an object) [0.410, 0.162, 0.490, 0.830] -> run/jog [0.150, 0.260, 0.396, 0.816] -> touch (an object) [0.106, 0.135, 0.264, 0.792]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.223, 0.16, 0.452, 0.803] location?", "sub_answer_type": "Action Trajectory" } ]

[ 900, 1772 ]

[ { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 914, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.006, 0.127, 0.14, 0.978] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1147, "options": [ "A. 5", "B. 8", "C. 4", "D. 3" ], "question": "How many people were performing the listen to (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1214, "options": [ "A. stir", "B. smoke", "C. walk and carry/hold (an object)", "D. carry/hold (an object) and smoke" ], "question": "What action is the person currently in the [0.285, 0.011, 0.985, 0.987] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1260, "options": [ "A. crawl [0.831, 0.154, 0.826, 0.824] -> talk to (e.g., self, a person, a group) [0.494, 0.274, 1.000, 0.730] -> write [0.819, 0.239, 0.973, 1.000]", "B. touch (an object) [0.689, 0.346, 1.000, 1.000] -> write [0.595, 0.364, 1.000, 0.724] -> talk to (e.g., self, a person, a group) [0.861, 0.263, 0.838, 0.738]", "C. touch (an object) [0.590, 0.107, 1.000, 0.697] -> talk to (e.g., self, a person, a group) [0.829, 0.410, 0.803, 0.914] -> dress/put on clothing [0.619, 0.209, 0.852, 0.954]", "D. touch (an object) [0.67, 0.247, 0.999, 0.87] -> write [0.67, 0.247, 0.999, 0.87] -> talk to (e.g., self, a person, a group) [0.67, 0.247, 0.999, 0.87]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.67, 0.247, 0.999, 0.87] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1270, "options": [ "A. 2", "B. 3", "C. 4", "D. 5" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1285, "options": [ "A. [0.202, 0.151, 0.599, 1.0]", "B. [0.331, 0.0, 0.53, 0.701]", "C. [0.255, 0.0, 0.818, 0.999]", "D. [0.398, 0.0, 1.0, 0.753]" ], "question": "What location in the frame is the person currently in the [0.0, 0.204, 0.378, 0.99] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1305, "options": [ "A. 5", "B. 7", "C. 3", "D. 4" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1377, "options": [ "A. Yes", "B. No" ], "question": "Are there still 4 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1407, "options": [ "A. 3", "B. 6", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: eat?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1464, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1528, "options": [ "A. 8", "B. 5", "C. 4", "D. 3" ], "question": "How many people were performing the eat in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1607, "options": [ "A. 4", "B. 6", "C. 3", "D. 5" ], "question": "How many people were performing the listen to (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1663, "options": [ "A. [0.087, 0.017, 0.884, 0.989]", "B. [0.009, 0.0, 0.561, 0.75]", "C. [0.223, 0.354, 0.777, 1.0]", "D. [0.0, 0.071, 0.596, 0.994]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1663, "options": [ "A. 2", "B. 3", "C. 4", "D. 6" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1671, "options": [ "A. carry/hold (an object) [0.005, 0.232, 0.628, 0.977] -> touch (an object) [0.005, 0.232, 0.628, 0.977] -> talk to (e.g., self, a person, a group) [0.005, 0.232, 0.628, 0.977]", "B. touch (an object) [0.000, 0.336, 0.698, 0.945] -> cut [0.000, 0.346, 0.822, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.187, 0.514, 1.000]", "C. touch (an object) [0.000, 0.163, 0.710, 1.000] -> sail boat [0.117, 0.372, 0.724, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.254, 0.570, 0.874]", "D. talk to (e.g., self, a person, a group) [0.000, 0.099, 0.597, 1.000] -> touch (an object) [0.000, 0.231, 0.806, 0.856] -> fall down [0.186, 0.062, 0.649, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.005, 0.232, 0.628, 0.977] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1673, "options": [ "A. touch (an object) [0.004, 0.209, 0.644, 0.966] -> talk to (e.g., self, a person, a group) [0.004, 0.209, 0.644, 0.966] -> carry/hold (an object) [0.026, 0.197, 0.658, 0.997]", "B. talk to (e.g., self, a person, a group) [0.000, 0.366, 0.448, 0.861] -> take a photo [0.000, 0.314, 0.510, 0.935] -> carry/hold (an object) [0.124, 0.253, 0.673, 0.994]", "C. talk to (e.g., self, a person, a group) [0.141, 0.272, 0.530, 1.000] -> talk to (e.g., self, a person, a group) [0.066, 0.149, 0.516, 1.000] -> touch (an object) [0.000, 0.069, 0.591, 0.968]", "D. give/serve (an object) to (a person) [0.000, 0.068, 0.464, 0.879] -> touch (an object) [0.092, 0.058, 0.832, 0.878] -> carry/hold (an object) [0.000, 0.073, 0.686, 0.864]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.026, 0.197, 0.658, 0.997] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1686, "options": [ "A. kick (an object) [0.000, 0.204, 0.656, 0.886] -> talk to (e.g., self, a person, a group) [0.117, 0.107, 0.787, 0.959] -> carry/hold (an object) [0.320, 0.274, 0.688, 0.874]", "B. talk to (e.g., self, a person, a group) [0.312, 0.266, 0.613, 0.994] -> dig [0.028, 0.246, 0.775, 0.998] -> touch (an object) [0.022, 0.085, 0.818, 0.970]", "C. carry/hold (an object) [0.13, 0.183, 0.651, 0.897] -> touch (an object) [0.13, 0.183, 0.651, 0.897] -> talk to (e.g., self, a person, a group) [0.13, 0.183, 0.651, 0.897]", "D. carry/hold (an object) [0.000, 0.045, 0.609, 1.000] -> eat [0.046, 0.014, 0.698, 0.903] -> touch (an object) [0.121, 0.253, 0.827, 0.843]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.13, 0.183, 0.651, 0.897] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1697, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.057, 0.242, 0.864] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1726, "options": [ "A. catch (an object) [0.127, 0.193, 0.904, 1.000] -> talk to (e.g., self, a person, a group) [0.153, 0.529, 0.725, 1.000] -> carry/hold (an object) [0.397, 0.472, 0.657, 1.000]", "B. carry/hold (an object) [0.223, 0.35, 0.738, 0.941] -> touch (an object) [0.223, 0.35, 0.738, 0.941] -> talk to (e.g., self, a person, a group) [0.223, 0.35, 0.738, 0.941]", "C. carry/hold (an object) [0.309, 0.331, 0.687, 0.942] -> pull (an object) [0.370, 0.451, 0.571, 1.000] -> talk to (e.g., self, a person, a group) [0.259, 0.440, 0.677, 0.825]", "D. dig [0.132, 0.228, 0.824, 0.966] -> talk to (e.g., self, a person, a group) [0.063, 0.451, 0.562, 0.875] -> touch (an object) [0.088, 0.344, 0.627, 0.751]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.223, 0.35, 0.738, 0.941] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1783, "options": [ "A. sit, carry/hold (an object), touch (an object), listen to (a person), watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. walk, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.221, 0.394, 0.364, 0.998] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 900, 1783 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 982, "options": [ "A. stand, watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object)", "D. stand, carry/hold (an object), listen to (a person)" ], "question": "What action is the person at the location [0.301, 0.268, 0.435, 0.598] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1085, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1085, "options": [ "A. 7", "B. 5", "C. 3", "D. 4" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1087, "options": [ "A. close (e.g., a door, a box)", "B. carry/hold (an object) and talk to (e.g., self, a person, a group)", "C. sing to (e.g., self, a person, a group)", "D. walk and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.122, 0.01, 0.617, 0.993] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1144, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.205, 0.357, 0.439, 0.982] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1165, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. sit, listen to (a person)" ], "question": "What action is the person at the location [0.1, 0.115, 0.446, 0.979] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1171, "options": [ "A. [0.305, 0.273, 0.436, 0.803]", "B. [0.079, 0.303, 0.295, 0.99]", "C. [0.482, 0.218, 0.599, 0.836]", "D. [0.043, 0.566, 0.156, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1320, "options": [ "A. [0.3, 0.336, 0.799, 0.458]", "B. [0.017, 0.213, 0.254, 0.967]", "C. [0.818, 0.101, 0.924, 0.745]", "D. [0.583, 0.147, 0.847, 0.757]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1344, "options": [ "A. stand, touch (an object), talk to (e.g., self, a person, a group)", "B. stand, listen to (a person)", "C. stand, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.027, 0.008, 0.574, 0.967] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1348, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1485, "options": [ "A. [0.343, 0.402, 0.867, 1.0]", "B. [0.405, 0.125, 0.811, 0.947]", "C. [0.347, 0.0, 1.0, 0.744]", "D. [0.093, 0.014, 0.583, 0.973]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1506, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1667, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. walk, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, watch (a person)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.302, 0.284, 0.634, 0.989] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1669, "options": [ "A. [0.604, 0.305, 0.679, 1.0]", "B. [0.436, 0.0, 0.887, 0.801]", "C. [0.349, 0.033, 0.741, 0.964]", "D. [0.35, 0.312, 0.433, 0.776]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1792, "options": [ "A. 4", "B. 3", "C. 6", "D. 5" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1792 ]

[ { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 946, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.431, 0.624, 0.514, 0.944] location in the current frame performing the close (e.g., a door, a box)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 946, "options": [ "A. walk [0.269, 0.575, 0.657, 1.000] -> sing to (e.g., self, a person, a group) [0.361, 0.677, 0.442, 0.896] -> close (e.g., a door, a box) [0.263, 0.820, 0.602, 0.776]", "B. carry/hold (an object) [0.391, 0.638, 0.447, 0.913] -> close (e.g., a door, a box) [0.391, 0.638, 0.447, 0.913] -> walk [0.431, 0.624, 0.514, 0.944]", "C. jump/leap [0.487, 0.485, 0.481, 0.759] -> walk [0.536, 0.742, 0.478, 0.916] -> carry/hold (an object) [0.316, 0.512, 0.403, 0.840]", "D. walk [0.613, 0.448, 0.481, 0.932] -> close (e.g., a door, a box) [0.588, 0.544, 0.580, 0.897] -> talk to (e.g., self, a person, a group) [0.424, 0.500, 0.516, 0.863]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.431, 0.624, 0.514, 0.944] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 990, "options": [ "A. touch (an object) -> crouch/kneel", "B. touch (an object) -> exit", "C. kick (an object) -> lie/sleep", "D. touch (an object) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 992, "options": [ "A. [0.542, 0.15, 0.828, 0.698]", "B. [0.061, 0.088, 0.932, 0.88]", "C. [0.044, 0.0, 0.656, 0.802]", "D. [0.0, 0.152, 0.65, 1.0]" ], "question": "Where is the person currently performing the touch (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1110, "options": [ "A. talk to (e.g., self, a person, a group) [0.317, 0.484, 0.200, 0.809] -> walk [0.273, 0.652, 0.591, 0.836] -> press [0.515, 0.683, 0.233, 0.745]", "B. extract [0.218, 0.778, 0.400, 1.000] -> carry/hold (an object) [0.440, 0.556, 0.459, 0.782] -> talk to (e.g., self, a person, a group) [0.331, 0.671, 0.271, 1.000]", "C. walk [0.318, 0.625, 0.392, 0.931] -> carry/hold (an object) [0.318, 0.625, 0.392, 0.931] -> talk to (e.g., self, a person, a group) [0.318, 0.625, 0.392, 0.931]", "D. carry/hold (an object) [0.122, 0.752, 0.238, 0.874] -> talk to (e.g., self, a person, a group) [0.122, 0.652, 0.280, 0.772] -> cut [0.331, 0.425, 0.456, 0.799]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.318, 0.625, 0.392, 0.931] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1122, "options": [ "A. catch (an object) -> bend/bow (at the waist)", "B. bend/bow (at the waist) -> lift (a person)", "C. press -> lift (a person)", "D. lift/pick up -> bend/bow (at the waist)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1132, "options": [ "A. 4", "B. 6", "C. 3", "D. 2" ], "question": "How many people in the current frame are performing the action: watch (e.g., TV)?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1196, "options": [ "A. 3", "B. 4", "C. 6", "D. 5" ], "question": "How many people were performing the sit in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1217, "options": [ "A. carry/hold (an object) [0.167, 0.000, 0.636, 1.000] -> fight/hit (a person) [0.151, 0.089, 0.762, 1.000] -> push (another person) [0.477, 0.326, 0.696, 1.000]", "B. walk [0.309, 0.145, 0.745, 0.973] -> carry/hold (an object) [0.309, 0.145, 0.745, 0.973] -> push (another person) [0.309, 0.145, 0.745, 0.973]", "C. stir [0.359, 0.073, 0.645, 1.000] -> walk [0.148, 0.101, 0.905, 1.000] -> push (another person) [0.382, 0.000, 0.579, 0.824]", "D. walk [0.219, 0.153, 0.870, 1.000] -> press [0.215, 0.297, 0.565, 0.903] -> carry/hold (an object) [0.432, 0.000, 0.675, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.309, 0.145, 0.745, 0.973] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1302, "options": [ "A. hit (an object) -> lie/sleep", "B. lie/sleep -> catch (an object)", "C. carry/hold (an object) -> paint", "D. carry/hold (an object) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1306, "options": [ "A. lie/sleep [0.341, 0.544, 0.626, 0.596] -> exit [0.293, 0.389, 0.517, 0.586] -> carry/hold (an object) [0.165, 0.258, 0.266, 0.697]", "B. shovel [0.367, 0.248, 0.596, 0.634] -> carry/hold (an object) [0.000, 0.501, 0.597, 0.959] -> lie/sleep [0.004, 0.204, 0.605, 0.902]", "C. talk to (e.g., self, a person, a group) [0.251, 0.398, 0.311, 0.910] -> open (e.g., a window, a car door) [0.020, 0.497, 0.527, 0.858] -> lie/sleep [0.372, 0.279, 0.322, 0.856]", "D. carry/hold (an object) [0.175, 0.364, 0.462, 0.769] -> talk to (e.g., self, a person, a group) [0.175, 0.364, 0.462, 0.769] -> lie/sleep [0.175, 0.364, 0.462, 0.769]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.175, 0.364, 0.462, 0.769] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1311, "options": [ "A. 4", "B. 6", "C. 8", "D. 3" ], "question": "How many people were performing the sit in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1333, "options": [ "A. [0.232, 0.034, 0.979, 0.977]", "B. [0.0, 0.106, 0.68, 1.0]", "C. [0.0, 0.0, 0.97, 0.752]", "D. [0.703, 0.011, 0.988, 0.527]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1372, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.463, 0.309, 0.697, 0.983] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1388, "options": [ "A. 4", "B. 3", "C. 6", "D. 5" ], "question": "How many people were performing the sit in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1405, "options": [ "A. carry/hold (an object) -> crouch/kneel", "B. crouch/kneel -> fall down", "C. fishing -> crouch/kneel", "D. crouch/kneel -> swim" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1482, "options": [ "A. [0.218, 0.0, 0.709, 0.785]", "B. [0.558, 0.711, 0.591, 0.888]", "C. [0.0, 0.019, 0.679, 0.995]", "D. [0.0, 0.21, 0.952, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1543, "options": [ "A. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> hug (a person)", "C. take (an object) from (a person) -> bend/bow (at the waist)", "D. talk to (e.g., self, a person, a group) -> fall down" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1550, "options": [ "A. carry/hold (an object) [0.545, 0.200, 0.664, 0.717] -> work on a computer [0.811, 0.392, 0.889, 0.959] -> grab (a person) [0.896, 0.180, 0.619, 0.923]", "B. walk [0.706, 0.324, 0.781, 0.817] -> carry/hold (an object) [0.706, 0.324, 0.781, 0.817] -> grab (a person) [0.706, 0.324, 0.781, 0.817]", "C. grab (a person) [0.832, 0.141, 0.791, 1.000] -> extract [0.739, 0.294, 0.976, 0.729] -> walk [0.667, 0.211, 0.880, 0.884]", "D. grab (a person) [0.659, 0.255, 0.642, 0.687] -> walk [0.629, 0.272, 0.644, 0.786] -> crawl [0.566, 0.146, 0.908, 0.862]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.706, 0.324, 0.781, 0.817] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1603, "options": [ "A. sit, ride (e.g., a bike, a car, a horse), listen to (a person)", "B. sit, ride (e.g., a bike, a car, a horse), talk to (e.g., self, a person, a group)", "C. stand, watch (a person)", "D. sit, carry/hold (an object)" ], "question": "What action is the person at the location [0.068, 0.154, 0.823, 0.973] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1710, "options": [ "A. swim", "B. play musical instrument", "C. crouch/kneel", "D. bend/bow (at the waist)" ], "question": "What action is the person currently in the [0.261, 0.309, 0.377, 0.688] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1711, "options": [ "A. sit, ride (e.g., a bike, a car, a horse), listen to (a person)", "B. sit, ride (e.g., a bike, a car, a horse)", "C. stand, listen to (a person), watch (a person)", "D. sit, play musical instrument, watch (a person)" ], "question": "What action is the person at the location [0.067, 0.309, 0.22, 0.737] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1730, "options": [ "A. 4", "B. 5", "C. 6", "D. 3" ], "question": "How many people were performing the walk in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1764, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1764 ]

[ { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 918, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 920, "options": [ "A. get up -> carry/hold (an object)", "B. carry/hold (an object) -> put down", "C. hand clap -> carry/hold (an object)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 928, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> answer phone", "C. talk to (e.g., self, a person, a group) -> cut", "D. talk to (e.g., self, a person, a group) -> clink glass" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 984, "options": [ "A. [0.024, 0.373, 0.68, 0.961]", "B. [0.204, 0.092, 0.691, 0.972]", "C. [0.15, 0.0, 0.431, 1.0]", "D. [0.451, 0.0, 0.784, 1.0]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1002, "options": [ "A. [0.355, 0.156, 0.95, 0.993]", "B. [0.281, 0.441, 0.692, 0.824]", "C. [0.376, 0.0, 0.836, 0.802]", "D. [0.081, 0.052, 0.473, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1085, "options": [ "A. [0.477, 0.147, 0.611, 0.482]", "B. [0.152, 0.048, 0.523, 0.972]", "C. [0.0, 0.294, 0.486, 1.0]", "D. [0.047, 0.0, 0.307, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1280, "options": [ "A. [0.0, 0.0, 0.689, 0.821]", "B. [0.096, 0.0, 1.0, 0.713]", "C. [0.341, 0.006, 0.969, 0.99]", "D. [0.133, 0.015, 0.859, 0.993]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1361, "options": [ "A. walk", "B. lie/sleep", "C. row boat", "D. talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.175, 0.0, 0.99, 0.987] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1440, "options": [ "A. enter -> carry/hold (an object)", "B. walk -> carry/hold (an object)", "C. carry/hold (an object) -> play with kids", "D. carry/hold (an object) -> cut" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1469, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.031, 0.051, 0.679, 0.936] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1471, "options": [ "A. [0.295, 0.379, 1.0, 1.0]", "B. [0.565, 0.033, 0.667, 0.684]", "C. [0.297, 0.08, 0.946, 0.952]", "D. [0.018, 0.0, 0.52, 0.975]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1481, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. walk, talk to (e.g., self, a person, a group)", "C. sit, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.0, 0.007, 0.621, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1500, "options": [ "A. carry/hold (an object) -> turn (e.g., a screwdriver)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> walk", "D. talk to (e.g., self, a person, a group) -> crawl" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1557, "options": [ "A. walk -> brush teeth", "B. talk to (e.g., self, a person, a group) -> paint", "C. talk to (e.g., self, a person, a group) -> walk", "D. talk to (e.g., self, a person, a group) -> write" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1565, "options": [ "A. stand, talk to (e.g., self, a person, a group)", "B. stand, carry/hold (an object)", "C. sit, listen to (a person), watch (a person)", "D. walk, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.13, 0.07, 0.405, 0.986] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1576, "options": [ "A. [0.057, 0.0, 0.482, 0.742]", "B. [0.151, 0.07, 0.429, 0.982]", "C. [0.0, 0.151, 0.146, 0.752]", "D. [0.27, 0.113, 0.956, 0.913]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1617, "options": [ "A. [0.694, 0.232, 0.727, 0.771]", "B. [0.751, 0.369, 0.89, 1.0]", "C. [0.464, 0.242, 0.708, 0.975]", "D. [0.014, 0.026, 0.465, 0.99]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1648, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1667, "options": [ "A. walk, talk to (e.g., self, a person, a group)", "B. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "C. stand, talk to (e.g., self, a person, a group)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.002, 0.001, 0.637, 0.972] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 900, 1667 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 911, "options": [ "A. bend/bow (at the waist), watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. stand, watch (a person)", "D. stand, listen to (a person)" ], "question": "What action is the person at the location [0.142, 0.134, 0.605, 0.981] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1123, "options": [ "A. [0.445, 0.3, 0.731, 0.848]", "B. [0.134, 0.085, 0.623, 0.984]", "C. [0.366, 0.139, 0.667, 0.698]", "D. [0.0, 0.0, 0.456, 1.0]" ], "question": "Where was the person currently performing the kiss (a person) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1170, "options": [ "A. carry/hold (an object) and talk to (e.g., self, a person, a group)", "B. walk", "C. cut", "D. shoot" ], "question": "What action is the person currently in the [0.586, 0.146, 0.873, 0.964] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1214, "options": [ "A. talk to (e.g., self, a person, a group) [0.455, 0.000, 0.509, 0.724] -> carry/hold (an object) [0.382, 0.036, 0.728, 0.737] -> throw [0.373, 0.212, 0.659, 0.480]", "B. play with pets [0.584, 0.000, 0.667, 0.486] -> talk to (e.g., self, a person, a group) [0.431, 0.068, 0.792, 0.636] -> bend/bow (at the waist) [0.477, 0.097, 0.707, 0.521]", "C. bend/bow (at the waist) [0.4, 0.053, 0.703, 0.554] -> carry/hold (an object) [0.4, 0.053, 0.703, 0.554] -> talk to (e.g., self, a person, a group) [0.4, 0.053, 0.703, 0.554]", "D. text on/look at a cellphone [0.270, 0.175, 0.786, 0.456] -> carry/hold (an object) [0.263, 0.000, 0.672, 0.393] -> talk to (e.g., self, a person, a group) [0.311, 0.035, 0.886, 0.615]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.4, 0.053, 0.703, 0.554] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1221, "options": [ "A. [0.538, 0.0, 0.889, 0.822]", "B. [0.131, 0.199, 0.397, 0.986]", "C. [0.032, 0.119, 0.516, 0.714]", "D. [0.299, 0.071, 0.645, 0.981]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1232, "options": [ "A. [0.467, 0.057, 0.915, 0.428]", "B. [0.385, 0.071, 0.77, 0.655]", "C. [0.491, 0.247, 0.94, 0.904]", "D. [0.305, 0.07, 0.642, 0.976]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1244, "options": [ "A. [0.588, 0.454, 0.853, 0.999]", "B. [0.812, 0.218, 0.96, 1.0]", "C. [0.356, 0.304, 0.627, 0.824]", "D. [0.859, 0.642, 0.861, 1.0]" ], "question": "Where was the person currently performing the eat in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1253, "options": [ "A. carry/hold (an object) [0.348, 0.000, 0.654, 0.535] -> play with kids [0.541, 0.000, 0.699, 0.773] -> talk to (e.g., self, a person, a group) [0.446, 0.000, 0.811, 0.462]", "B. bend/bow (at the waist) [0.399, 0.111, 0.716, 0.658] -> carry/hold (an object) [0.399, 0.111, 0.716, 0.658] -> talk to (e.g., self, a person, a group) [0.399, 0.111, 0.716, 0.658]", "C. carry/hold (an object) [0.514, 0.058, 0.766, 0.676] -> lie/sleep [0.562, 0.158, 0.532, 0.730] -> bend/bow (at the waist) [0.335, 0.209, 0.629, 0.728]", "D. talk to (e.g., self, a person, a group) [0.510, 0.302, 0.584, 0.719] -> carry/hold (an object) [0.451, 0.000, 0.733, 0.481] -> cut [0.534, 0.000, 0.888, 0.698]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.399, 0.111, 0.716, 0.658] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1352, "options": [ "A. Yes", "B. No" ], "question": "Are there still 5 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1406, "options": [ "A. carry/hold (an object) [0.761, 0.370, 0.739, 1.000] -> walk [0.707, 0.224, 0.664, 0.981] -> exit [0.742, 0.217, 0.723, 1.000]", "B. walk [0.588, 0.219, 0.774, 0.988] -> carry/hold (an object) [0.588, 0.219, 0.774, 0.988] -> talk to (e.g., self, a person, a group) [0.588, 0.219, 0.774, 0.988]", "C. hug (a person) [0.523, 0.033, 0.949, 0.837] -> walk [0.580, 0.292, 0.590, 1.000] -> talk to (e.g., self, a person, a group) [0.590, 0.364, 0.914, 0.810]", "D. carry/hold (an object) [0.768, 0.295, 0.668, 1.000] -> walk [0.451, 0.325, 0.896, 1.000] -> smoke [0.569, 0.076, 0.949, 0.865]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.588, 0.219, 0.774, 0.988] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1431, "options": [ "A. stand, carry/hold (an object), watch (a person)", "B. sit, listen to (a person), watch (a person)", "C. stand, listen to (a person)", "D. walk, listen to (a person)" ], "question": "What action is the person at the location [0.128, 0.083, 0.47, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1541, "options": [ "A. [0.173, 0.049, 0.553, 0.96]", "B. [0.225, 0.0, 0.818, 0.798]", "C. [0.378, 0.299, 0.638, 0.979]", "D. [0.0, 0.241, 0.375, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1556, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1607, "options": [ "A. carry/hold (an object) [0.000, 0.000, 0.215, 0.808] -> walk [0.311, 0.233, 0.399, 0.786] -> turn (e.g., a screwdriver) [0.174, 0.026, 0.245, 0.916]", "B. answer phone [0.275, 0.021, 0.449, 0.995] -> walk [0.049, 0.019, 0.385, 0.879] -> carry/hold (an object) [0.324, 0.000, 0.346, 0.790]", "C. walk [0.057, 0.236, 0.251, 0.774] -> carry/hold (an object) [0.071, 0.236, 0.527, 1.000] -> hand wave [0.254, 0.245, 0.273, 1.000]", "D. walk [0.129, 0.122, 0.365, 0.969] -> carry/hold (an object) [0.129, 0.122, 0.365, 0.969] -> talk to (e.g., self, a person, a group) [0.129, 0.122, 0.365, 0.969]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.129, 0.122, 0.365, 0.969] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1649, "options": [ "A. stand, carry/hold (an object), talk to (e.g., self, a person, a group)", "B. stand, carry/hold (an object), listen to (a person), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.263, 0.101, 0.508, 0.973] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1665, "options": [ "A. [0.292, 0.0, 0.717, 0.93]", "B. [0.155, 0.058, 0.516, 0.998]", "C. [0.0, 0.288, 0.346, 0.831]", "D. [0.406, 0.464, 0.506, 0.826]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1703, "options": [ "A. [0.745, 0.35, 1.0, 0.719]", "B. [0.461, 0.447, 0.87, 0.993]", "C. [0.29, 0.74, 1.0, 1.0]", "D. [0.135, 0.166, 0.329, 0.917]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1756, "options": [ "A. exit [0.034, 0.310, 0.438, 0.871] -> hug (a person) [0.031, 0.055, 0.705, 1.000] -> lie/sleep [0.313, 0.225, 0.528, 1.000]", "B. talk to (e.g., self, a person, a group) [0.155, 0.236, 0.555, 0.974] -> lie/sleep [0.155, 0.236, 0.555, 0.974] -> hug (a person) [0.124, 0.24, 0.547, 0.99]", "C. talk to (e.g., self, a person, a group) [0.060, 0.307, 0.654, 0.871] -> exit [0.199, 0.392, 0.484, 1.000] -> lie/sleep [0.189, 0.324, 0.442, 1.000]", "D. talk to (e.g., self, a person, a group) [0.000, 0.391, 0.561, 0.960] -> lie/sleep [0.159, 0.122, 0.660, 1.000] -> cook [0.215, 0.163, 0.745, 0.994]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.124, 0.24, 0.547, 0.99] location?", "sub_answer_type": "Action Trajectory" } ]

[ 900, 1756 ]

[ { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 916, "options": [ "A. carry/hold (an object) [0.326, 0.109, 0.488, 0.910] -> hand clap [0.101, 0.433, 0.651, 0.837] -> talk to (e.g., self, a person, a group) [0.414, 0.111, 0.880, 0.882]", "B. sing to (e.g., self, a person, a group) [0.156, 0.268, 0.711, 0.996] -> close (e.g., a door, a box) [0.190, 0.205, 0.800, 0.832] -> carry/hold (an object) [0.203, 0.073, 0.713, 0.916]", "C. talk to (e.g., self, a person, a group) [0.283, 0.198, 0.856, 0.97] -> carry/hold (an object) [0.245, 0.25, 0.606, 0.877] -> close (e.g., a door, a box) [0.245, 0.25, 0.606, 0.877]", "D. carry/hold (an object) [0.310, 0.277, 0.713, 0.908] -> kiss (a person) [0.051, 0.445, 0.492, 1.000] -> talk to (e.g., self, a person, a group) [0.394, 0.354, 0.691, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.245, 0.25, 0.606, 0.877] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 917, "options": [ "A. open (e.g., a window, a car door) -> talk to (e.g., self, a person, a group) -> carry/hold (an object) -> grab (a person)", "B. carry/hold (an object) -> open (e.g., a window, a car door) -> talk to (e.g., self, a person, a group) -> close (e.g., a door, a box)", "C. close (e.g., a door, a box) -> martial art -> open (e.g., a window, a car door) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> carry/hold (an object) -> grab (a person) -> close (e.g., a door, a box)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 992, "options": [ "A. talk to (e.g., self, a person, a group) [0.144, 0.000, 0.377, 0.851] -> give/serve (an object) to (a person) [0.045, 0.238, 0.503, 0.874] -> turn (e.g., a screwdriver) [0.062, 0.220, 0.447, 1.000]", "B. talk to (e.g., self, a person, a group) [0.008, 0.000, 0.461, 1.000] -> throw [0.000, 0.000, 0.442, 0.776] -> give/serve (an object) to (a person) [0.000, 0.000, 0.457, 0.903]", "C. carry/hold (an object) [0.024, 0.079, 0.522, 0.96] -> give/serve (an object) to (a person) [0.024, 0.079, 0.522, 0.96] -> talk to (e.g., self, a person, a group) [0.024, 0.079, 0.522, 0.96]", "D. dress/put on clothing [0.126, 0.154, 0.374, 0.779] -> carry/hold (an object) [0.000, 0.245, 0.697, 0.924] -> give/serve (an object) to (a person) [0.109, 0.005, 0.569, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.024, 0.079, 0.522, 0.96] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 997, "options": [ "A. read [0.480, 0.354, 0.915, 0.768] -> talk to (e.g., self, a person, a group) [0.263, 0.213, 1.000, 1.000] -> text on/look at a cellphone [0.280, 0.138, 0.875, 0.783]", "B. talk to (e.g., self, a person, a group) [0.613, 0.182, 1.000, 0.875] -> carry/hold (an object) [0.553, 0.120, 1.000, 1.000] -> eat [0.363, 0.035, 1.000, 0.840]", "C. enter [0.224, 0.342, 0.934, 1.000] -> read [0.376, 0.218, 1.000, 0.906] -> carry/hold (an object) [0.346, 0.364, 1.000, 0.953]", "D. carry/hold (an object) [0.414, 0.206, 0.961, 0.869] -> read [0.414, 0.206, 0.961, 0.869] -> talk to (e.g., self, a person, a group) [0.414, 0.206, 0.961, 0.869]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.414, 0.206, 0.961, 0.869] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1001, "options": [ "A. carry/hold (an object) [0.432, 0.127, 1.000, 0.712] -> talk to (e.g., self, a person, a group) [0.287, 0.374, 1.000, 0.866] -> crouch/kneel [0.344, 0.357, 1.000, 0.926]", "B. carry/hold (an object) [0.317, 0.025, 0.957, 0.753] -> kick (an object) [0.364, 0.046, 0.900, 1.000] -> read [0.605, 0.365, 1.000, 0.964]", "C. read [0.426, 0.187, 0.953, 0.856] -> talk to (e.g., self, a person, a group) [0.426, 0.187, 0.953, 0.856] -> carry/hold (an object) [0.394, 0.199, 0.943, 0.85]", "D. sail boat [0.294, 0.182, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.588, 0.187, 1.000, 0.995] -> carry/hold (an object) [0.518, 0.328, 1.000, 0.726]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.394, 0.199, 0.943, 0.85] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1012, "options": [ "A. [0.0, 0.216, 0.047, 0.813]", "B. [0.025, 0.083, 0.335, 0.966]", "C. [0.414, 0.206, 0.961, 0.869]", "D. [0.321, 0.0, 0.16, 0.714]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1022, "options": [ "A. [0.249, 0.281, 0.289, 0.414]", "B. [0.548, 0.024, 0.492, 0.714]", "C. [0.302, 0.241, 0.757, 1.0]", "D. [0.212, 0.394, 0.518, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1036, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1044, "options": [ "A. lift/pick up -> read", "B. talk to (e.g., self, a person, a group) -> extract", "C. take (an object) from (a person) -> read", "D. talk to (e.g., self, a person, a group) -> read" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1053, "options": [ "A. push (another person) -> touch (an object)", "B. touch (an object) -> carry/hold (an object)", "C. smoke -> touch (an object)", "D. touch (an object) -> enter" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1058, "options": [ "A. talk to (e.g., self, a person, a group) -> pull (an object)", "B. carry/hold (an object) -> enter", "C. talk to (e.g., self, a person, a group) -> brush teeth", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1164, "options": [ "A. [0.579, 0.624, 0.698, 1.0]", "B. [0.724, 0.425, 0.84, 0.958]", "C. [0.426, 0.43, 0.688, 0.94]", "D. [0.334, 0.08, 0.856, 0.996]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1235, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.394, 0.326, 0.712, 0.979] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1278, "options": [ "A. [0.032, 0.105, 0.129, 0.781]", "B. [0.136, 0.027, 0.921, 0.999]", "C. [0.069, 0.123, 0.405, 0.947]", "D. [0.0, 0.0, 0.596, 0.749]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1299, "options": [ "A. [0.029, 0.153, 0.545, 0.986]", "B. [0.139, 0.405, 0.839, 1.0]", "C. [0.0, 0.115, 0.642, 0.695]", "D. [0.509, 0.167, 0.819, 0.983]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1308, "options": [ "A. stand, carry/hold (an object), watch (a person)", "B. stand, carry/hold (an object), ride (e.g., a bike, a car, a horse)", "C. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.016, 0.079, 0.44, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1366, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.638, 0.233, 0.76, 0.699] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1442, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.457, 0.154, 0.598, 0.646] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1443, "options": [ "A. 8", "B. 4", "C. 2", "D. 3" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1528, "options": [ "A. sail boat [0.486, 0.327, 0.875, 1.000] -> carry/hold (an object) [0.322, 0.104, 0.729, 1.000] -> talk to (e.g., self, a person, a group) [0.567, 0.270, 0.867, 1.000]", "B. talk to (e.g., self, a person, a group) [0.415, 0.172, 0.934, 0.99] -> carry/hold (an object) [0.327, 0.186, 0.716, 0.985] -> give/serve (an object) to (a person) [0.327, 0.186, 0.716, 0.985]", "C. give/serve (an object) to (a person) [0.504, 0.348, 0.800, 0.863] -> carry/hold (an object) [0.389, 0.130, 0.795, 1.000] -> dig [0.489, 0.321, 1.000, 0.885]", "D. cut [0.327, 0.017, 0.810, 0.866] -> carry/hold (an object) [0.492, 0.138, 0.775, 0.842] -> give/serve (an object) to (a person) [0.409, 0.169, 0.828, 0.900]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.327, 0.186, 0.716, 0.985] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1560, "options": [ "A. 3", "B. 5", "C. 4", "D. 7" ], "question": "How many people were performing the watch (a person) in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1592, "options": [ "A. kick (an object)", "B. carry/hold (an object) and ride (e.g., a bike, a car, a horse)", "C. carry/hold (an object)", "D. swim" ], "question": "What action is the person currently in the [0.188, 0.108, 0.701, 0.998] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1593, "options": [ "A. answer phone -> hit (an object) -> carry/hold (an object)", "B. walk -> carry/hold (an object) -> hit (an object)", "C. hit (an object) -> carry/hold (an object) -> crouch/kneel", "D. carry/hold (an object) -> hit (an object) -> grab (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1627, "options": [ "A. 3", "B. 9", "C. 7", "D. 4" ], "question": "How many people were performing the watch (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1678, "options": [ "A. Yes", "B. No" ], "question": "Are there still 4 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1715, "options": [ "A. 3", "B. 5", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" } ]

[ 900, 1715 ]

[ { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 971, "options": [ "A. [0.258, 0.0, 0.684, 0.687]", "B. [0.26, 0.0, 1.0, 0.902]", "C. [0.027, 0.112, 0.983, 0.965]", "D. [0.234, 0.324, 0.532, 0.947]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 988, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1038, "options": [ "A. [0.335, 0.094, 0.705, 0.972]", "B. [0.123, 0.0, 0.451, 0.793]", "C. [0.536, 0.0, 0.848, 0.751]", "D. [0.039, 0.196, 0.517, 0.966]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1071, "options": [ "A. [0.305, 0.027, 0.964, 0.969]", "B. [0.258, 0.363, 1.0, 1.0]", "C. [0.193, 0.489, 0.616, 1.0]", "D. [0.107, 0.202, 0.912, 0.963]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1188, "options": [ "A. put down [0.482, 0.046, 0.693, 0.872] -> bend/bow (at the waist) [0.518, 0.024, 0.551, 0.695] -> carry/hold (an object) [0.331, 0.305, 0.525, 0.700]", "B. carry/hold (an object) [0.509, 0.073, 0.653, 0.622] -> bend/bow (at the waist) [0.624, 0.248, 0.586, 0.773] -> push (an object) [0.543, 0.079, 0.665, 0.958]", "C. carry/hold (an object) [0.424, 0.156, 0.649, 0.797] -> smoke [0.424, 0.156, 0.649, 0.797] -> bend/bow (at the waist) [0.425, 0.179, 0.655, 0.797]", "D. bend/bow (at the waist) [0.620, 0.364, 0.514, 0.747] -> close (e.g., a door, a box) [0.486, 0.268, 0.838, 0.688] -> carry/hold (an object) [0.375, 0.299, 0.754, 0.800]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.425, 0.179, 0.655, 0.797] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1200, "options": [ "A. ride (e.g., a bike, a car, a horse) [0.603, 0.232, 0.962, 0.881] -> touch (an object) [0.613, 0.368, 0.964, 1.000] -> talk to (e.g., self, a person, a group) [0.662, 0.167, 0.884, 1.000]", "B. carry/hold (an object) [0.674, 0.215, 0.992, 0.850] -> touch (an object) [0.725, 0.388, 0.886, 0.988] -> sing to (e.g., self, a person, a group) [0.404, 0.479, 1.000, 0.954]", "C. talk to (e.g., self, a person, a group) [0.634, 0.197, 1.000, 1.000] -> touch (an object) [0.490, 0.423, 0.989, 0.995] -> crouch/kneel [0.588, 0.195, 1.000, 1.000]", "D. talk to (e.g., self, a person, a group) [0.575, 0.283, 0.984, 0.962] -> carry/hold (an object) [0.577, 0.286, 0.981, 0.962] -> touch (an object) [0.577, 0.286, 0.981, 0.962]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.577, 0.286, 0.981, 0.962] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1206, "options": [ "A. [0.0, 0.293, 0.41, 0.998]", "B. [0.244, 0.176, 0.552, 0.853]", "C. [0.0, 0.154, 0.436, 0.727]", "D. [0.169, 0.344, 0.614, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1214, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.023, 0.243, 0.735, 0.96] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1231, "options": [ "A. talk to (e.g., self, a person, a group) [0.01, 0.259, 0.539, 0.978] -> smoke [0.027, 0.293, 0.529, 0.987] -> touch (an object) [0.027, 0.293, 0.529, 0.987]", "B. talk to (e.g., self, a person, a group) [0.185, 0.424, 0.439, 1.000] -> smoke [0.000, 0.444, 0.652, 0.939] -> shoot [0.175, 0.176, 0.433, 0.863]", "C. smoke [0.000, 0.316, 0.633, 1.000] -> lie/sleep [0.121, 0.180, 0.466, 0.869] -> touch (an object) [0.000, 0.229, 0.461, 0.916]", "D. smoke [0.156, 0.213, 0.701, 0.996] -> crawl [0.165, 0.219, 0.565, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.334, 0.533, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.027, 0.293, 0.529, 0.987] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1254, "options": [ "A. [0.492, 0.479, 0.901, 1.0]", "B. [0.566, 0.148, 0.911, 0.971]", "C. [0.148, 0.584, 0.489, 0.978]", "D. [0.268, 0.314, 0.659, 0.978]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1333, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, touch (an object), listen to (a person), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. stand, watch (a person)" ], "question": "What action is the person at the location [0.019, 0.2, 0.181, 0.613] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1407, "options": [ "A. talk to (e.g., self, a person, a group) [0.386, 0.309, 0.739, 1.000] -> touch (an object) [0.375, 0.327, 0.792, 0.866] -> run/jog [0.326, 0.362, 0.814, 0.850]", "B. grab (a person) [0.173, 0.414, 0.696, 0.757] -> touch (an object) [0.237, 0.235, 0.704, 0.742] -> carry/hold (an object) [0.191, 0.499, 0.712, 0.807]", "C. carry/hold (an object) [0.495, 0.230, 0.832, 0.896] -> shoot [0.271, 0.191, 0.508, 0.782] -> talk to (e.g., self, a person, a group) [0.352, 0.207, 0.880, 0.985]", "D. carry/hold (an object) [0.35, 0.39, 0.704, 0.912] -> touch (an object) [0.35, 0.39, 0.704, 0.912] -> talk to (e.g., self, a person, a group) [0.35, 0.39, 0.704, 0.912]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.35, 0.39, 0.704, 0.912] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1417, "options": [ "A. [0.17, 0.471, 0.572, 1.0]", "B. [0.622, 0.212, 0.631, 0.686]", "C. [0.518, 0.24, 0.982, 0.955]", "D. [0.346, 0.376, 0.715, 0.889]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1420, "options": [ "A. sit, carry/hold (an object), touch (an object), listen to (a person)", "B. sit, touch (an object), watch (a person)", "C. sit, smoke, touch (an object), listen to (a person)", "D. sit, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.003, 0.099, 0.14, 0.959] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1545, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1586, "options": [ "A. walk, listen to (a person), watch (a person)", "B. sit, carry/hold (an object), touch (an object), listen to (a person)", "C. sit, touch (an object), watch (a person)", "D. sit, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.008, 0.139, 0.605, 0.962] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1594, "options": [ "A. play musical instrument [0.214, 0.242, 0.641, 0.941] -> carry/hold (an object) [0.002, 0.024, 0.409, 1.000] -> talk to (e.g., self, a person, a group) [0.147, 0.218, 0.573, 1.000]", "B. touch (an object) [0.000, 0.294, 0.525, 1.000] -> carry/hold (an object) [0.134, 0.142, 0.585, 0.873] -> play with pets [0.022, 0.167, 0.689, 0.863]", "C. talk to (e.g., self, a person, a group) [0.029, 0.102, 0.552, 0.953] -> carry/hold (an object) [0.014, 0.129, 0.565, 0.95] -> touch (an object) [0.014, 0.129, 0.565, 0.95]", "D. carry/hold (an object) [0.000, 0.270, 0.730, 1.000] -> talk to (e.g., self, a person, a group) [0.032, 0.092, 0.458, 0.830] -> hand wave [0.000, 0.074, 0.708, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.014, 0.129, 0.565, 0.95] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1727, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1781, "options": [ "A. dig", "B. touch (an object)", "C. drink", "D. walk" ], "question": "What action is the person currently in the [0.59, 0.138, 0.902, 0.961] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1782, "options": [ "A. [0.014, 0.129, 0.565, 0.95]", "B. [0.556, 0.058, 0.732, 1.0]", "C. [0.709, 0.25, 0.881, 0.946]", "D. [0.691, 0.463, 0.624, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1782 ]

[ { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 907, "options": [ "A. dress/put on clothing -> carry/hold (an object)", "B. carry/hold (an object) -> play with kids", "C. answer phone -> carry/hold (an object)", "D. bend/bow (at the waist) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 961, "options": [ "A. [0.426, 0.311, 0.985, 0.982]", "B. [0.208, 0.071, 0.462, 0.58]", "C. [0.423, 0.134, 0.749, 0.869]", "D. [0.572, 0.114, 0.957, 0.869]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 973, "options": [ "A. walk [0.399, 0.228, 1.000, 1.000] -> exit [0.611, 0.128, 1.000, 0.689] -> talk to (e.g., self, a person, a group) [0.676, 0.000, 0.847, 0.797]", "B. fishing [0.508, 0.049, 0.877, 0.913] -> walk [0.641, 0.313, 1.000, 0.946] -> carry/hold (an object) [0.462, 0.000, 1.000, 1.000]", "C. walk [0.54, 0.155, 0.975, 0.85] -> carry/hold (an object) [0.54, 0.155, 0.975, 0.85] -> talk to (e.g., self, a person, a group) [0.54, 0.155, 0.975, 0.85]", "D. carry/hold (an object) [0.673, 0.223, 0.780, 0.737] -> put down [0.658, 0.317, 1.000, 0.804] -> talk to (e.g., self, a person, a group) [0.516, 0.000, 0.996, 0.823]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.54, 0.155, 0.975, 0.85] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 983, "options": [ "A. carry/hold (an object) [0.473, 0.138, 0.909, 0.803] -> hug (a person) [0.788, 0.099, 1.000, 0.776] -> talk to (e.g., self, a person, a group) [0.287, 0.020, 0.944, 0.833]", "B. carry/hold (an object) [0.37, 0.139, 0.791, 0.88] -> talk to (e.g., self, a person, a group) [0.37, 0.139, 0.791, 0.88] -> walk [0.678, 0.124, 0.993, 0.859]", "C. walk [0.645, 0.155, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.493, 0.098, 0.800, 1.000] -> fall down [0.322, 0.051, 0.769, 0.706]", "D. talk to (e.g., self, a person, a group) [0.466, 0.000, 0.692, 0.703] -> text on/look at a cellphone [0.496, 0.174, 0.794, 0.763] -> carry/hold (an object) [0.321, 0.286, 0.599, 0.945]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.678, 0.124, 0.993, 0.859] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1045, "options": [ "A. fall down [0.000, 0.096, 0.376, 0.750] -> carry/hold (an object) [0.000, 0.094, 0.534, 0.866] -> talk to (e.g., self, a person, a group) [0.071, 0.220, 0.569, 0.765]", "B. talk to (e.g., self, a person, a group) [0.08, 0.133, 0.503, 0.772] -> carry/hold (an object) [0.046, 0.142, 0.51, 0.788] -> touch (an object) [0.046, 0.142, 0.51, 0.788]", "C. touch (an object) [0.000, 0.040, 0.361, 0.711] -> stir [0.000, 0.083, 0.447, 0.624] -> carry/hold (an object) [0.086, 0.000, 0.489, 0.675]", "D. talk to (e.g., self, a person, a group) [0.152, 0.294, 0.456, 0.718] -> bend/bow (at the waist) [0.128, 0.029, 0.583, 0.888] -> touch (an object) [0.201, 0.170, 0.645, 0.645]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.046, 0.142, 0.51, 0.788] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1047, "options": [ "A. [0.75, 0.278, 0.979, 0.868]", "B. [0.417, 0.0, 0.806, 0.629]", "C. [0.184, 0.431, 0.298, 0.718]", "D. [0.156, 0.141, 0.519, 0.772]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1064, "options": [ "A. Yes", "B. No" ], "question": "Are there still 4 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1161, "options": [ "A. 4", "B. 2", "C. 3", "D. 5" ], "question": "How many people in the current frame are performing the action: dance?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1164, "options": [ "A. dance [0.527, 0.161, 0.620, 0.783] -> write [0.280, 0.192, 0.572, 0.759] -> talk to (e.g., self, a person, a group) [0.502, 0.228, 0.505, 0.882]", "B. kick (an object) [0.174, 0.436, 0.506, 1.000] -> carry/hold (an object) [0.419, 0.137, 0.608, 0.873] -> talk to (e.g., self, a person, a group) [0.334, 0.266, 0.850, 0.853]", "C. carry/hold (an object) [0.374, 0.32, 0.696, 0.888] -> dance [0.374, 0.32, 0.696, 0.888] -> talk to (e.g., self, a person, a group) [0.374, 0.32, 0.696, 0.888]", "D. talk to (e.g., self, a person, a group) [0.421, 0.446, 0.778, 0.745] -> chop [0.445, 0.353, 0.794, 1.000] -> carry/hold (an object) [0.290, 0.199, 0.884, 0.707]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.374, 0.32, 0.696, 0.888] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1207, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.27, 0.199, 0.505, 0.866] location in the current frame performing the touch (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1227, "options": [ "A. ride (e.g., a bike, a car, a horse) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> talk to (e.g., self, a person, a group)", "C. chop -> ride (e.g., a bike, a car, a horse)", "D. ride (e.g., a bike, a car, a horse) -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1262, "options": [ "A. [0.431, 0.272, 0.748, 0.869]", "B. [0.544, 0.089, 0.95, 1.0]", "C. [0.189, 0.197, 0.508, 0.866]", "D. [0.444, 0.126, 0.627, 0.599]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1294, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.427, 0.241, 0.71, 0.856] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1311, "options": [ "A. enter -> carry/hold (an object) -> close (e.g., a door, a box)", "B. close (e.g., a door, a box) -> walk -> carry/hold (an object)", "C. press -> carry/hold (an object) -> close (e.g., a door, a box)", "D. close (e.g., a door, a box) -> bend/bow (at the waist) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Future Prediction", "middle_frame_timestamp": 1316, "options": [ "A. fight/hit (a person)", "B. walk and carry/hold (an object)", "C. carry/hold (an object) and talk to (e.g., self, a person, a group)", "D. sing to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.225, 0.201, 0.441, 0.869] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1319, "options": [ "A. talk to (e.g., self, a person, a group) [0.407, 0.106, 0.327, 0.795] -> push (an object) [0.208, 0.177, 0.327, 0.781] -> carry/hold (an object) [0.109, 0.282, 0.591, 0.717]", "B. carry/hold (an object) [0.248, 0.192, 0.455, 0.876] -> talk to (e.g., self, a person, a group) [0.248, 0.192, 0.455, 0.876] -> walk [0.238, 0.191, 0.414, 0.869]", "C. walk [0.177, 0.123, 0.540, 0.897] -> eat [0.244, 0.347, 0.309, 0.930] -> carry/hold (an object) [0.257, 0.104, 0.323, 0.748]", "D. carry/hold (an object) [0.391, 0.301, 0.600, 0.948] -> walk [0.383, 0.219, 0.543, 1.000] -> push (another person) [0.151, 0.057, 0.574, 0.997]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.238, 0.191, 0.414, 0.869] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1328, "options": [ "A. [0.597, 0.933, 0.717, 0.836]", "B. [0.72, 0.673, 0.912, 0.869]", "C. [0.0, 0.223, 0.332, 0.869]", "D. [0.809, 0.476, 0.624, 0.815]" ], "question": "What location in the frame is the person currently in the [0.416, 0.453, 0.621, 0.864] location likely to move to next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1336, "options": [ "A. dress/put on clothing -> run/jog", "B. talk to (e.g., self, a person, a group) -> run/jog", "C. run/jog -> sing to (e.g., self, a person, a group)", "D. run/jog -> write" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1363, "options": [ "A. [0.151, 0.194, 0.537, 0.777]", "B. [0.11, 0.506, 0.429, 0.831]", "C. [0.459, 0.163, 0.977, 0.913]", "D. [0.407, 0.364, 0.552, 0.866]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1461, "options": [ "A. 5", "B. 4", "C. 3", "D. 7" ], "question": "How many people were performing the stand in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1506, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1561, "options": [ "A. [0.465, 0.0, 0.969, 0.701]", "B. [0.257, 0.243, 0.823, 0.86]", "C. [0.172, 0.267, 0.496, 0.872]", "D. [0.19, 0.442, 0.913, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1641, "options": [ "A. 4", "B. 3", "C. 6", "D. 5" ], "question": "How many people were performing the listen to (a person) in the scene 240 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1692, "options": [ "A. [0.809, 0.298, 0.978, 0.719]", "B. [0.062, 0.234, 0.801, 0.851]", "C. [0.737, 0.593, 0.879, 0.756]", "D. [0.564, 0.311, 0.691, 0.774]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1703, "options": [ "A. 6", "B. 7", "C. 3", "D. 5" ], "question": "How many people were performing the stand in the scene 300 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1751, "options": [ "A. talk to (e.g., self, a person, a group) -> climb (e.g., a mountain) -> carry/hold (an object)", "B. talk to (e.g., self, a person, a group) -> carry/hold (an object) -> lift/pick up", "C. carry/hold (an object) -> read -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> dress/put on clothing -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1751 ]

[ { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1006, "options": [ "A. read [0.131, 0.473, 0.525, 0.845] -> talk to (e.g., self, a person, a group) [0.230, 0.242, 0.497, 1.000] -> touch (an object) [0.152, 0.265, 0.372, 0.704]", "B. talk to (e.g., self, a person, a group) [0.362, 0.278, 0.520, 0.876] -> swim [0.166, 0.185, 0.596, 0.968] -> read [0.394, 0.278, 0.426, 1.000]", "C. carry/hold (an object) [0.144, 0.286, 0.607, 0.810] -> sail boat [0.393, 0.128, 0.409, 1.000] -> read [0.076, 0.097, 0.495, 1.000]", "D. read [0.212, 0.285, 0.497, 0.9] -> talk to (e.g., self, a person, a group) [0.212, 0.285, 0.497, 0.9] -> carry/hold (an object) [0.206, 0.285, 0.483, 0.889]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.206, 0.285, 0.483, 0.889] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1009, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1016, "options": [ "A. take a photo [0.246, 0.131, 0.499, 0.933] -> walk [0.025, 0.303, 0.339, 0.835] -> read [0.285, 0.254, 0.391, 0.876]", "B. play board game [0.076, 0.286, 0.589, 0.966] -> carry/hold (an object) [0.284, 0.133, 0.354, 1.000] -> walk [0.176, 0.241, 0.578, 0.802]", "C. read [0.101, 0.176, 0.421, 0.986] -> walk [0.105, 0.137, 0.436, 0.998] -> carry/hold (an object) [0.105, 0.137, 0.436, 0.998]", "D. walk [0.171, 0.000, 0.464, 1.000] -> brush teeth [0.000, 0.092, 0.234, 0.840] -> carry/hold (an object) [0.076, 0.023, 0.280, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.105, 0.137, 0.436, 0.998] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1030, "options": [ "A. 4", "B. 2", "C. 3", "D. 1" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1031, "options": [ "A. sit, listen to (a person)", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. walk, carry/hold (an object), listen to (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.064, 0.118, 0.397, 0.99] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1034, "options": [ "A. kiss (a person) [0.000, 0.016, 0.455, 0.868] -> carry/hold (an object) [0.033, 0.203, 0.525, 0.957] -> read [0.000, 0.273, 0.363, 0.804] -> talk to (e.g., self, a person, a group) [0.067, 0.171, 0.555, 1.000]", "B. talk to (e.g., self, a person, a group) [0.017, 0.123, 0.368, 1.0] -> walk [0.0, 0.118, 0.381, 1.0] -> carry/hold (an object) [0.0, 0.118, 0.381, 1.0] -> read [0.0, 0.118, 0.381, 1.0]", "C. walk [0.000, 0.284, 0.548, 1.000] -> open (e.g., a window, a car door) [0.133, 0.164, 0.562, 0.892] -> talk to (e.g., self, a person, a group) [0.000, 0.023, 0.518, 1.000] -> carry/hold (an object) [0.000, 0.000, 0.573, 1.000]", "D. talk to (e.g., self, a person, a group) [0.000, 0.286, 0.496, 1.000] -> give/serve (an object) to (a person) [0.085, 0.000, 0.190, 1.000] -> read [0.000, 0.020, 0.393, 1.000] -> walk [0.007, 0.016, 0.494, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.0, 0.118, 0.381, 1.0] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1075, "options": [ "A. 4", "B. 1", "C. 3", "D. 2" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1125, "options": [ "A. walk -> carry/hold (an object)", "B. play with kids -> walk", "C. carry/hold (an object) -> ride (e.g., a bike, a car, a horse)", "D. carry/hold (an object) -> exit" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1164, "options": [ "A. [0.374, 0.115, 0.761, 0.872]", "B. [0.284, 0.354, 0.977, 1.0]", "C. [0.591, 0.0, 0.996, 0.716]", "D. [0.031, 0.014, 0.717, 0.998]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1268, "options": [ "A. give/serve (an object) to (a person) -> walk", "B. walk -> talk to (e.g., self, a person, a group)", "C. walk -> row boat", "D. walk -> throw" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1297, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1303, "options": [ "A. carry/hold (an object) -> watch (e.g., TV)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> answer phone", "D. talk to (e.g., self, a person, a group) -> cook" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1317, "options": [ "A. sail boat [0.543, 0.464, 1.000, 0.889] -> talk to (e.g., self, a person, a group) [0.499, 0.273, 0.826, 0.963] -> talk to (e.g., self, a person, a group) [0.498, 0.260, 0.997, 0.659]", "B. catch (an object) [0.549, 0.277, 1.000, 0.623] -> sail boat [0.745, 0.347, 0.853, 0.621] -> carry/hold (an object) [0.653, 0.371, 0.706, 0.658]", "C. catch (an object) [0.733, 0.550, 0.710, 0.978] -> talk to (e.g., self, a person, a group) [0.509, 0.239, 0.755, 0.646] -> carry/hold (an object) [0.576, 0.431, 0.683, 0.635]", "D. talk to (e.g., self, a person, a group) [0.591, 0.425, 0.877, 0.803] -> carry/hold (an object) [0.587, 0.428, 0.874, 0.809] -> sail boat [0.587, 0.428, 0.874, 0.809]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.587, 0.428, 0.874, 0.809] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1332, "options": [ "A. give/serve (an object) to (a person) [0.042, 0.000, 0.680, 0.887] -> talk to (e.g., self, a person, a group) [0.074, 0.155, 0.856, 1.000] -> cut [0.000, 0.223, 0.728, 1.000]", "B. give/serve (an object) to (a person) [0.087, 0.075, 0.486, 1.000] -> talk to (e.g., self, a person, a group) [0.184, 0.000, 0.703, 0.964] -> read [0.000, 0.078, 0.589, 0.876]", "C. row boat [0.150, 0.087, 0.604, 0.806] -> carry/hold (an object) [0.000, 0.066, 0.559, 0.868] -> talk to (e.g., self, a person, a group) [0.079, 0.077, 0.530, 0.991]", "D. carry/hold (an object) [0.018, 0.088, 0.665, 0.99] -> give/serve (an object) to (a person) [0.018, 0.088, 0.665, 0.99] -> talk to (e.g., self, a person, a group) [0.018, 0.088, 0.665, 0.99]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.018, 0.088, 0.665, 0.99] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1348, "options": [ "A. [0.094, 0.125, 0.891, 0.975]", "B. [0.296, 0.154, 0.999, 0.986]", "C. [0.475, 0.434, 1.0, 0.962]", "D. [0.196, 0.011, 1.0, 0.687]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1355, "options": [ "A. carry/hold (an object) -> play musical instrument", "B. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "D. talk to (e.g., self, a person, a group) -> play musical instrument" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1382, "options": [ "A. talk to (e.g., self, a person, a group) -> point to (an object)", "B. talk to (e.g., self, a person, a group) -> dress/put on clothing", "C. talk to (e.g., self, a person, a group) -> cut", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1444, "options": [ "A. sit, listen to (a person)", "B. walk, carry/hold (an object)", "C. sit, carry/hold (an object), listen to (a person), watch (a person)", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.011, 0.193, 0.296, 0.977] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1462, "options": [ "A. 2", "B. 3", "C. 1", "D. 4" ], "question": "What was the minimum number of people in the scene over the past 15 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1527, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.35, 0.12, 0.676, 0.608] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1598, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1614, "options": [ "A. play musical instrument", "B. smoke", "C. dig", "D. talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.316, 0.008, 1.0, 1.0] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1623, "options": [ "A. 2", "B. 1", "C. 3", "D. 4" ], "question": "What was the minimum number of people in the scene over the past 30 seconds?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1671, "options": [ "A. [0.327, 0.226, 0.656, 0.969]", "B. [0.25, 0.176, 0.905, 0.766]", "C. [0.0, 0.088, 0.644, 0.986]", "D. [0.0, 0.0, 0.369, 0.903]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1686, "options": [ "A. [0.279, 0.2, 0.703, 0.792]", "B. [0.555, 0.481, 0.698, 0.651]", "C. [0.045, 0.159, 0.618, 0.998]", "D. [0.259, 0.0, 0.961, 0.499]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1686 ]

[ { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 951, "options": [ "A. [0.027, 0.0, 0.835, 1.0]", "B. [0.829, 0.092, 0.815, 0.699]", "C. [0.673, 0.348, 0.94, 1.0]", "D. [0.555, 0.18, 0.843, 0.949]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1017, "options": [ "A. walk -> work on a computer", "B. carry/hold (an object) -> write", "C. walk -> kick (an object)", "D. walk -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1061, "options": [ "A. touch (an object) [0.066, 0.03, 0.961, 1.0] -> talk to (e.g., self, a person, a group) [0.066, 0.03, 0.961, 1.0] -> drink [0.605, 0.238, 0.959, 0.92]", "B. catch (an object) [0.071, 0.064, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.241, 0.000, 1.000, 0.857] -> drink [0.635, 0.370, 0.893, 0.897]", "C. touch (an object) [0.000, 0.000, 0.996, 0.842] -> run/jog [0.100, 0.000, 1.000, 0.806] -> drink [0.499, 0.328, 0.960, 0.791]", "D. lift (a person) [0.682, 0.064, 1.000, 0.977] -> touch (an object) [0.176, 0.000, 0.766, 0.864] -> talk to (e.g., self, a person, a group) [0.000, 0.000, 0.786, 0.929]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.605, 0.238, 0.959, 0.92] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1064, "options": [ "A. watch (e.g., TV) [0.588, 0.152, 0.822, 0.744] -> touch (an object) [0.673, 0.285, 0.784, 0.758] -> drink [0.387, 0.211, 0.832, 0.878]", "B. talk to (e.g., self, a person, a group) [0.742, 0.103, 0.866, 0.741] -> drink [0.715, 0.249, 0.975, 0.853] -> shovel [0.654, 0.175, 1.000, 0.887]", "C. touch (an object) [0.607, 0.146, 0.926, 0.964] -> swim [0.675, 0.308, 0.800, 0.744] -> drink [0.506, 0.466, 0.772, 0.970]", "D. touch (an object) [0.561, 0.252, 0.968, 0.94] -> talk to (e.g., self, a person, a group) [0.561, 0.252, 0.968, 0.94] -> drink [0.559, 0.287, 0.957, 0.929]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.559, 0.287, 0.957, 0.929] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1068, "options": [ "A. drink [0.572, 0.253, 0.969, 0.94] -> touch (an object) [0.572, 0.253, 0.969, 0.94] -> talk to (e.g., self, a person, a group) [0.572, 0.253, 0.969, 0.94]", "B. drink [0.735, 0.305, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.724, 0.249, 0.979, 0.880] -> turn (e.g., a screwdriver) [0.708, 0.299, 1.000, 0.909]", "C. drink [0.382, 0.248, 0.925, 1.000] -> touch (an object) [0.621, 0.376, 0.770, 1.000] -> hand wave [0.575, 0.220, 1.000, 0.922]", "D. talk to (e.g., self, a person, a group) [0.623, 0.216, 1.000, 1.000] -> touch (an object) [0.471, 0.164, 1.000, 0.870] -> give/serve (an object) to (a person) [0.513, 0.201, 0.937, 0.900]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.572, 0.253, 0.969, 0.94] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1072, "options": [ "A. touch (an object) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> clink glass", "C. push (an object) -> touch (an object)", "D. smoke -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1079, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1081, "options": [ "A. sit, answer phone", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, answer phone, carry/hold (an object), touch (an object), talk to (e.g., self, a person, a group)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.021, 0.207, 0.477, 0.983] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1136, "options": [ "A. sit, answer phone, talk to (e.g., self, a person, a group)", "B. sit, answer phone, touch (an object)", "C. sit, answer phone", "D. sit, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.003, 0.113, 0.455, 0.978] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1154, "options": [ "A. [0.0, 0.115, 0.492, 0.98]", "B. [0.563, 0.24, 0.684, 0.657]", "C. [0.0, 0.025, 0.729, 0.709]", "D. [0.039, 0.0, 0.368, 0.713]" ], "question": "Where was the person currently performing the answer phone in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1166, "options": [ "A. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "B. put down -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> hit (an object)", "D. hit (an object) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1183, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1250, "options": [ "A. [0.029, 0.185, 0.425, 0.981]", "B. [0.151, 0.0, 1.0, 1.0]", "C. [0.0, 0.021, 0.824, 0.993]", "D. [0.0, 0.0, 0.662, 0.728]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1274, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.001, 0.017, 0.877, 0.993] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1298, "options": [ "A. push (an object) -> touch (an object) -> get up -> talk to (e.g., self, a person, a group) -> bend/bow (at the waist)", "B. talk to (e.g., self, a person, a group) -> touch (an object) -> bend/bow (at the waist) -> get up -> lift/pick up", "C. carry/hold (an object) -> get up -> touch (an object) -> talk to (e.g., self, a person, a group) -> throw", "D. touch (an object) -> talk to (e.g., self, a person, a group) -> get up -> bend/bow (at the waist) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1343, "options": [ "A. talk to (e.g., self, a person, a group) [0.590, 0.182, 0.990, 0.843] -> fishing [0.557, 0.213, 1.000, 0.859] -> touch (an object) [0.526, 0.000, 0.763, 1.000]", "B. eat [0.490, 0.290, 1.000, 0.890] -> carry/hold (an object) [0.720, 0.150, 1.000, 0.776] -> touch (an object) [0.847, 0.268, 1.000, 0.925]", "C. carry/hold (an object) [0.688, 0.125, 0.935, 0.964] -> touch (an object) [0.688, 0.125, 0.935, 0.964] -> talk to (e.g., self, a person, a group) [0.688, 0.125, 0.935, 0.964]", "D. touch (an object) [0.749, 0.278, 0.964, 0.863] -> carry/hold (an object) [0.711, 0.179, 0.855, 0.967] -> chop [0.602, 0.187, 1.000, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.688, 0.125, 0.935, 0.964] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1346, "options": [ "A. sit, read, touch (an object)", "B. sit, touch (an object), listen to (a person), watch (a person)", "C. walk, carry/hold (an object)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.186, 0.475, 0.286, 0.867] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1353, "options": [ "A. talk to (e.g., self, a person, a group)", "B. walk", "C. play with pets", "D. work on a computer" ], "question": "What action is the person currently in the [0.029, 0.295, 0.414, 0.979] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1438, "options": [ "A. sit, answer phone", "B. sit, touch (an object), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person), watch (a person)", "D. sit, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.136, 0.306, 0.446, 0.783] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1440, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1450, "options": [ "A. climb (e.g., a mountain) [0.247, 0.143, 1.000, 0.795] -> read [0.473, 0.000, 0.786, 1.000] -> touch (an object) [0.255, 0.000, 0.847, 1.000]", "B. read [0.275, 0.109, 0.977, 0.953] -> touch (an object) [0.275, 0.109, 0.977, 0.953] -> talk to (e.g., self, a person, a group) [0.275, 0.109, 0.977, 0.953]", "C. close (e.g., a door, a box) [0.158, 0.084, 0.966, 0.952] -> read [0.078, 0.000, 1.000, 0.878] -> touch (an object) [0.139, 0.000, 0.904, 0.843]", "D. read [0.103, 0.256, 1.000, 1.000] -> read [0.297, 0.199, 0.838, 1.000] -> touch (an object) [0.077, 0.285, 0.918, 0.917]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.275, 0.109, 0.977, 0.953] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1592, "options": [ "A. [0.407, 0.0, 0.296, 0.463]", "B. [0.256, 0.0, 0.775, 0.579]", "C. [0.361, 0.162, 0.59, 0.758]", "D. [0.0, 0.115, 0.492, 0.98]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1658, "options": [ "A. sit, answer phone, touch (an object)", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, answer phone", "D. stand, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.078, 0.292, 0.325, 0.985] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1708, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1749, "options": [ "A. walk -> brush teeth", "B. walk -> work on a computer", "C. walk -> carry/hold (an object)", "D. walk -> answer phone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1781, "options": [ "A. [0.31, 0.082, 0.989, 0.972]", "B. [0.186, 0.377, 0.775, 1.0]", "C. [0.078, 0.136, 0.479, 0.978]", "D. [0.117, 0.0, 0.761, 0.813]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1781 ]

[ { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 909, "options": [ "A. carry/hold (an object) -> point to (an object)", "B. eat -> carry/hold (an object)", "C. hand clap -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 938, "options": [ "A. paint [0.230, 0.548, 0.619, 0.975] -> talk to (e.g., self, a person, a group) [0.048, 0.568, 0.316, 1.000] -> walk [0.185, 0.465, 0.360, 0.946]", "B. push (an object) [0.075, 0.400, 0.632, 1.000] -> talk to (e.g., self, a person, a group) [0.244, 0.480, 0.651, 0.824] -> walk [0.030, 0.550, 0.472, 0.988]", "C. carry/hold (an object) [0.158, 0.487, 0.456, 1.0] -> talk to (e.g., self, a person, a group) [0.158, 0.487, 0.456, 1.0] -> walk [0.14, 0.586, 0.451, 0.984]", "D. walk [0.122, 0.417, 0.597, 1.000] -> hug (a person) [0.251, 0.515, 0.596, 1.000] -> carry/hold (an object) [0.025, 0.591, 0.369, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.14, 0.586, 0.451, 0.984] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 957, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1007, "options": [ "A. 5", "B. 2", "C. 4", "D. 3" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1025, "options": [ "A. dress/put on clothing -> carry/hold (an object)", "B. touch (an object) -> carry/hold (an object)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> text on/look at a cellphone" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1045, "options": [ "A. 7", "B. 6", "C. 5", "D. 4" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1063, "options": [ "A. text on/look at a cellphone -> touch (an object) -> answer phone -> talk to (e.g., self, a person, a group)", "B. text on/look at a cellphone -> touch (an object) -> answer phone -> bend/bow (at the waist)", "C. touch (an object) -> talk to (e.g., self, a person, a group) -> hug (a person) -> text on/look at a cellphone", "D. answer phone -> touch (an object) -> push (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1106, "options": [ "A. 5", "B. 3", "C. 2", "D. 4" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1190, "options": [ "A. 5", "B. 6", "C. 3", "D. 4" ], "question": "How many people were performing the dance in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1225, "options": [ "A. 6", "B. 4", "C. 5", "D. 3" ], "question": "How many people were performing the watch (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1301, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1327, "options": [ "A. talk to (e.g., self, a person, a group) [0.559, 0.368, 0.537, 0.931] -> run/jog [0.484, 0.252, 0.724, 0.823] -> work on a computer [0.681, 0.165, 0.654, 1.000]", "B. give/serve (an object) to (a person) [0.727, 0.305, 0.732, 0.795] -> run/jog [0.416, 0.502, 0.701, 1.000] -> talk to (e.g., self, a person, a group) [0.558, 0.395, 0.764, 0.941]", "C. run/jog [0.734, 0.488, 0.626, 1.000] -> drive (e.g., a car, a truck) [0.455, 0.459, 0.806, 1.000] -> carry/hold (an object) [0.497, 0.345, 0.530, 1.000]", "D. run/jog [0.544, 0.347, 0.718, 0.983] -> carry/hold (an object) [0.544, 0.347, 0.718, 0.983] -> talk to (e.g., self, a person, a group) [0.544, 0.347, 0.718, 0.983]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.544, 0.347, 0.718, 0.983] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1378, "options": [ "A. 3", "B. 6", "C. 4", "D. 2" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1413, "options": [ "A. 7", "B. 3", "C. 5", "D. 6" ], "question": "How many people were performing the watch (a person) in the scene 300 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1497, "options": [ "A. climb (e.g., a mountain)", "B. fall down", "C. run/jog", "D. talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.154, 0.059, 0.778, 0.982] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1499, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1545, "options": [ "A. walk -> eat", "B. carry/hold (an object) -> kiss (a person)", "C. walk -> carry/hold (an object)", "D. walk -> turn (e.g., a screwdriver)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1563, "options": [ "A. fishing [0.303, 0.000, 0.941, 1.000] -> walk [0.583, 0.000, 0.606, 1.000] -> talk to (e.g., self, a person, a group) [0.434, 0.084, 0.957, 0.978]", "B. carry/hold (an object) [0.5, 0.07, 0.804, 0.975] -> talk to (e.g., self, a person, a group) [0.5, 0.07, 0.804, 0.975] -> walk [0.408, 0.078, 0.754, 0.99]", "C. walk [0.384, 0.266, 0.890, 1.000] -> talk to (e.g., self, a person, a group) [0.554, 0.236, 0.627, 0.878] -> stir [0.445, 0.258, 0.868, 0.889]", "D. enter [0.428, 0.099, 0.885, 0.912] -> walk [0.542, 0.125, 0.732, 1.000] -> carry/hold (an object) [0.551, 0.124, 0.920, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.408, 0.078, 0.754, 0.99] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1566, "options": [ "A. [0.221, 0.168, 0.321, 0.579]", "B. [0.418, 0.133, 0.686, 0.7]", "C. [0.232, 0.0, 0.597, 1.0]", "D. [0.52, 0.055, 0.869, 0.984]" ], "question": "Where was the person currently performing the grab (a person) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1571, "options": [ "A. talk to (e.g., self, a person, a group) -> kick (an object)", "B. hug (a person) -> talk to (e.g., self, a person, a group)", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. talk to (e.g., self, a person, a group) -> shovel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1606, "options": [ "A. [0.254, 0.076, 0.888, 0.99]", "B. [0.086, 0.257, 0.676, 0.744]", "C. [0.272, 0.354, 0.72, 1.0]", "D. [0.667, 0.438, 0.79, 0.702]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1719, "options": [ "A. walk [0.243, 0.288, 0.537, 0.992] -> grab (a person) [0.243, 0.288, 0.537, 0.992] -> talk to (e.g., self, a person, a group) [0.243, 0.288, 0.537, 0.992]", "B. play with pets [0.324, 0.343, 0.570, 1.000] -> walk [0.339, 0.390, 0.636, 0.857] -> talk to (e.g., self, a person, a group) [0.401, 0.265, 0.349, 1.000]", "C. talk to (e.g., self, a person, a group) [0.206, 0.458, 0.733, 1.000] -> carry/hold (an object) [0.329, 0.124, 0.571, 1.000] -> walk [0.428, 0.126, 0.382, 1.000]", "D. grab (a person) [0.312, 0.162, 0.659, 0.968] -> take (an object) from (a person) [0.276, 0.390, 0.538, 0.814] -> walk [0.080, 0.276, 0.447, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.243, 0.288, 0.537, 0.992] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1773, "options": [ "A. bend/bow (at the waist) [0.322, 0.502, 0.582, 0.641] -> push (another person) [0.285, 0.530, 0.610, 0.573] -> carry/hold (an object) [0.610, 0.609, 0.642, 0.560]", "B. lift/pick up [0.535, 0.689, 0.571, 0.747] -> dig [0.659, 0.663, 0.476, 0.604] -> bend/bow (at the waist) [0.326, 0.715, 0.684, 0.696]", "C. lift/pick up [0.342, 0.519, 0.561, 0.648] -> carry/hold (an object) [0.428, 0.544, 0.719, 0.594] -> martial art [0.497, 0.720, 0.391, 0.574]", "D. lift/pick up [0.458, 0.547, 0.526, 0.689] -> bend/bow (at the waist) [0.496, 0.563, 0.584, 0.675] -> carry/hold (an object) [0.496, 0.563, 0.584, 0.675]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.496, 0.563, 0.584, 0.675] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1793, "options": [ "A. stand, listen to (a person), watch (a person)", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. stand, carry/hold (an object)", "D. walk, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.241, 0.245, 0.453, 0.999] in the image performing?", "sub_answer_type": "Action Location" } ]

[ 900, 1793 ]

[ { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 919, "options": [ "A. [0.24, 0.219, 1.0, 1.0]", "B. [0.783, 0.245, 1.0, 0.719]", "C. [0.521, 0.059, 0.962, 0.983]", "D. [0.113, 0.482, 0.315, 0.946]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 978, "options": [ "A. [0.521, 0.059, 0.962, 0.983]", "B. [0.478, 0.297, 0.571, 0.673]", "C. [0.464, 0.525, 0.695, 0.845]", "D. [0.296, 0.163, 0.423, 0.896]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1012, "options": [ "A. [0.319, 0.267, 0.499, 0.5]", "B. [0.402, 0.322, 1.0, 0.906]", "C. [0.16, 0.057, 0.992, 0.996]", "D. [0.0, 0.315, 0.982, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1023, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1030, "options": [ "A. text on/look at a cellphone -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> clink glass", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "D. run/jog -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1043, "options": [ "A. [0.155, 0.326, 0.659, 0.963]", "B. [0.617, 0.09, 0.879, 0.932]", "C. [0.347, 0.196, 0.886, 0.569]", "D. [0.491, 0.312, 0.617, 0.774]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1046, "options": [ "A. open (e.g., a window, a car door) -> bend/bow (at the waist) -> carry/hold (an object)", "B. bend/bow (at the waist) -> throw -> open (e.g., a window, a car door)", "C. carry/hold (an object) -> crawl -> bend/bow (at the waist)", "D. open (e.g., a window, a car door) -> bend/bow (at the waist) -> sail boat" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1212, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.364, 0.194, 0.725, 0.996] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1229, "options": [ "A. [0.779, 0.229, 0.975, 0.98]", "B. [0.461, 0.226, 0.674, 0.944]", "C. [0.588, 0.082, 0.695, 0.791]", "D. [0.291, 0.0, 0.522, 0.785]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1308, "options": [ "A. sit, listen to (a person), talk to (e.g., self, a person, a group), watch (a person)", "B. sit, talk to (e.g., self, a person, a group), watch (a person)", "C. sit, talk to (e.g., self, a person, a group)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.028, 0.105, 0.488, 0.887] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1325, "options": [ "A. carry/hold (an object) -> cook", "B. eat -> extract", "C. swim -> eat", "D. carry/hold (an object) -> eat" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1367, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.657, 0.111, 0.839, 0.937] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1447, "options": [ "A. [0.181, 0.444, 0.287, 1.0]", "B. [0.236, 0.188, 0.495, 0.987]", "C. [0.386, 0.166, 0.98, 0.97]", "D. [0.408, 0.0, 0.685, 0.858]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1550, "options": [ "A. carry/hold (an object) and talk to (e.g., self, a person, a group)", "B. get up", "C. open (e.g., a window, a car door)", "D. pull (an object)" ], "question": "What action is the person currently in the [0.135, 0.306, 0.642, 0.972] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1552, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. sit, talk to (e.g., self, a person, a group)", "C. sit, listen to (a person), watch (a person)", "D. sit, lift/pick up, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.202, 0.429, 0.596, 0.98] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1638, "options": [ "A. touch (an object) -> carry/hold (an object)", "B. fishing -> carry/hold (an object)", "C. carry/hold (an object) -> cut", "D. carry/hold (an object) -> touch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1688, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1734, "options": [ "A. sit, listen to (a person), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. sit, touch (an object), listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.072, 0.311, 0.454, 0.978] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1758, "options": [ "A. [0.287, 0.025, 1.0, 1.0]", "B. [0.516, 0.053, 1.0, 0.776]", "C. [0.563, 0.307, 1.0, 0.965]", "D. [0.0, 0.067, 0.927, 0.988]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1788, "options": [ "A. stir -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> enter", "C. bend/bow (at the waist) -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1788 ]

[ { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 948, "options": [ "A. [0.172, 0.051, 0.704, 0.818]", "B. [0.328, 0.0, 0.85, 0.793]", "C. [0.386, 0.482, 0.56, 0.995]", "D. [0.393, 0.066, 0.976, 0.983]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 961, "options": [ "A. talk to (e.g., self, a person, a group) -> pull (an object)", "B. talk to (e.g., self, a person, a group) -> walk", "C. walk -> brush teeth", "D. push (another person) -> walk" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 970, "options": [ "A. [0.334, 0.334, 0.863, 1.0]", "B. [0.544, 0.343, 0.616, 0.591]", "C. [0.0, 0.0, 0.483, 0.724]", "D. [0.191, 0.072, 0.59, 0.983]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 974, "options": [ "A. cook -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "C. row boat -> talk to (e.g., self, a person, a group)", "D. carry/hold (an object) -> chop" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1053, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1091, "options": [ "A. walk", "B. jump/leap", "C. hand clap", "D. carry/hold (an object)" ], "question": "What action is the person currently in the [0.197, 0.441, 0.305, 0.96] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1200, "options": [ "A. [0.109, 0.141, 0.471, 0.977]", "B. [0.114, 0.0, 0.872, 0.717]", "C. [0.21, 0.429, 1.0, 1.0]", "D. [0.046, 0.138, 0.929, 0.942]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1221, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1239, "options": [ "A. bend/bow (at the waist) -> write", "B. martial art -> bend/bow (at the waist)", "C. bend/bow (at the waist) -> kiss (a person)", "D. dress/put on clothing -> kiss (a person)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1246, "options": [ "A. carry/hold (an object) [0.438, 0.558, 0.503, 0.718] -> crouch/kneel [0.438, 0.558, 0.503, 0.718] -> fishing [0.438, 0.558, 0.503, 0.718]", "B. fishing [0.307, 0.731, 0.580, 0.625] -> touch (an object) [0.550, 0.502, 0.592, 0.840] -> carry/hold (an object) [0.311, 0.483, 0.315, 0.650]", "C. carry/hold (an object) [0.587, 0.741, 0.375, 0.585] -> crouch/kneel [0.370, 0.725, 0.575, 0.797] -> read [0.253, 0.684, 0.691, 0.658]", "D. carry/hold (an object) [0.399, 0.378, 0.634, 0.631] -> crouch/kneel [0.557, 0.562, 0.323, 0.857] -> cut [0.302, 0.535, 0.662, 0.639]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.438, 0.558, 0.503, 0.718] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1269, "options": [ "A. carry/hold (an object) [0.401, 0.117, 0.657, 0.977] -> talk to (e.g., self, a person, a group) [0.401, 0.117, 0.657, 0.977] -> walk [0.408, 0.173, 0.705, 0.988]", "B. carry/hold (an object) [0.289, 0.138, 0.807, 0.850] -> talk to (e.g., self, a person, a group) [0.552, 0.000, 0.622, 0.857] -> fall down [0.577, 0.233, 0.526, 0.801]", "C. walk [0.389, 0.223, 0.669, 1.000] -> carry/hold (an object) [0.449, 0.000, 0.611, 0.786] -> shoot [0.394, 0.268, 0.775, 0.876]", "D. walk [0.411, 0.267, 0.748, 0.908] -> kiss (a person) [0.363, 0.166, 0.597, 1.000] -> carry/hold (an object) [0.417, 0.274, 0.641, 0.978]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.408, 0.173, 0.705, 0.988] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1271, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. play musical instrument -> carry/hold (an object)", "C. talk to (e.g., self, a person, a group) -> shoot", "D. carry/hold (an object) -> catch (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1276, "options": [ "A. fall down -> crouch/kneel", "B. work on a computer -> carry/hold (an object)", "C. carry/hold (an object) -> crouch/kneel", "D. carry/hold (an object) -> shovel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1356, "options": [ "A. [0.0, 0.175, 0.564, 0.692]", "B. [0.0, 0.438, 0.104, 1.0]", "C. [0.161, 0.02, 0.707, 0.977]", "D. [0.003, 0.184, 0.275, 0.98]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1356, "options": [ "A. stand, carry/hold (an object), listen to (a person), watch (a person)", "B. stand, carry/hold (an object)", "C. stand, carry/hold (an object), cook, watch (a person)", "D. dance, watch (a person)" ], "question": "What action is the person at the location [0.394, 0.234, 0.588, 0.853] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1415, "options": [ "A. jump/leap [0.325, 0.263, 0.659, 0.834] -> talk to (e.g., self, a person, a group) [0.505, 0.078, 0.540, 1.000] -> walk [0.398, 0.193, 0.611, 1.000] -> grab (a person) [0.070, 0.312, 0.763, 0.890]", "B. talk to (e.g., self, a person, a group) [0.306, 0.205, 0.625, 0.993] -> walk [0.256, 0.182, 0.607, 0.991] -> carry/hold (an object) [0.256, 0.182, 0.607, 0.991] -> grab (a person) [0.256, 0.182, 0.607, 0.991]", "C. talk to (e.g., self, a person, a group) [0.407, 0.224, 0.625, 1.000] -> grab (a person) [0.307, 0.021, 0.608, 1.000] -> walk [0.315, 0.358, 0.602, 1.000] -> swim [0.102, 0.095, 0.762, 1.000]", "D. grab (a person) [0.331, 0.102, 0.793, 1.000] -> smoke [0.214, 0.198, 0.764, 1.000] -> carry/hold (an object) [0.141, 0.361, 0.737, 1.000] -> walk [0.399, 0.275, 0.760, 0.848]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.256, 0.182, 0.607, 0.991] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1547, "options": [ "A. 6", "B. 4", "C. 3", "D. 5" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1553, "options": [ "A. Yes", "B. No" ], "question": "Are there still 3 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1573, "options": [ "A. sit, carry/hold (an object), eat", "B. stand, talk to (e.g., self, a person, a group), watch (a person)", "C. dance, watch (a person)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.051, 0.409, 0.302, 0.94] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1612, "options": [ "A. 4", "B. 7", "C. 5", "D. 3" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1681, "options": [ "A. carry/hold (an object) [0.329, 0.122, 0.423, 0.923] -> swim [0.397, 0.332, 0.406, 1.000] -> walk [0.248, 0.270, 0.601, 1.000]", "B. walk [0.348, 0.201, 0.558, 1.000] -> lift/pick up [0.457, 0.095, 0.628, 1.000] -> talk to (e.g., self, a person, a group) [0.351, 0.111, 0.411, 1.000]", "C. kick (a person) [0.443, 0.384, 0.545, 0.855] -> carry/hold (an object) [0.408, 0.017, 0.761, 1.000] -> walk [0.484, 0.000, 0.496, 0.852]", "D. walk [0.361, 0.194, 0.595, 0.987] -> carry/hold (an object) [0.361, 0.194, 0.595, 0.987] -> talk to (e.g., self, a person, a group) [0.361, 0.194, 0.595, 0.987]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.361, 0.194, 0.595, 0.987] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1685, "options": [ "A. walk [0.036, 0.438, 0.612, 1.000] -> dress/put on clothing [0.000, 0.524, 0.723, 1.000] -> carry/hold (an object) [0.184, 0.525, 0.393, 1.000]", "B. enter [0.002, 0.349, 0.482, 0.801] -> walk [0.214, 0.416, 0.556, 0.900] -> carry/hold (an object) [0.000, 0.339, 0.416, 0.855]", "C. walk [0.029, 0.368, 0.539, 0.995] -> carry/hold (an object) [0.029, 0.368, 0.539, 0.995] -> talk to (e.g., self, a person, a group) [0.029, 0.368, 0.539, 0.995]", "D. talk to (e.g., self, a person, a group) [0.156, 0.427, 0.593, 0.834] -> walk [0.082, 0.354, 0.622, 0.943] -> hand clap [0.039, 0.303, 0.506, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.029, 0.368, 0.539, 0.995] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1693, "options": [ "A. 2", "B. 4", "C. 5", "D. 3" ], "question": "How many people in the current frame are performing the action: carry/hold (an object)?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1709, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.365, 0.343, 0.947, 0.999] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1718, "options": [ "A. [0.0, 0.009, 0.557, 0.987]", "B. [0.294, 0.259, 0.542, 0.982]", "C. [0.061, 0.0, 0.759, 0.721]", "D. [0.032, 0.0, 0.272, 0.814]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1718 ]

[ { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 977, "options": [ "A. carry/hold (an object) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> sing to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> kick (a person)", "D. talk to (e.g., self, a person, a group) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 988, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1026, "options": [ "A. lie/sleep -> climb (e.g., a mountain)", "B. hit (an object) -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> lie/sleep", "D. run/jog -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1059, "options": [ "A. [0.484, 0.186, 0.73, 0.872]", "B. [0.515, 0.064, 0.427, 0.74]", "C. [0.26, 0.318, 0.421, 0.983]", "D. [0.251, 0.046, 0.153, 1.0]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1068, "options": [ "A. [0.014, 0.0, 0.768, 0.987]", "B. [0.0, 0.261, 0.477, 0.788]", "C. [0.107, 0.258, 0.908, 1.0]", "D. [0.223, 0.039, 0.935, 0.892]" ], "question": "Where is the person currently performing the touch (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1129, "options": [ "A. walk [0.541, 0.404, 0.666, 0.895] -> carry/hold (an object) [0.541, 0.404, 0.666, 0.895] -> talk to (e.g., self, a person, a group) [0.541, 0.404, 0.666, 0.895]", "B. talk to (e.g., self, a person, a group) [0.444, 0.495, 0.571, 0.727] -> sail boat [0.392, 0.588, 0.780, 0.808] -> carry/hold (an object) [0.737, 0.592, 0.830, 1.000]", "C. walk [0.681, 0.331, 0.811, 0.948] -> fight/hit (a person) [0.383, 0.246, 0.625, 0.817] -> carry/hold (an object) [0.513, 0.282, 0.684, 0.876]", "D. carry/hold (an object) [0.490, 0.411, 0.752, 0.942] -> exit [0.734, 0.596, 0.609, 1.000] -> walk [0.642, 0.387, 0.614, 0.938]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.541, 0.404, 0.666, 0.895] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1136, "options": [ "A. play with kids [0.320, 0.456, 0.465, 0.853] -> carry/hold (an object) [0.313, 0.201, 0.785, 1.000] -> talk to (e.g., self, a person, a group) [0.481, 0.439, 0.520, 0.808]", "B. bend/bow (at the waist) [0.453, 0.321, 0.553, 0.797] -> walk [0.487, 0.451, 0.799, 1.000] -> carry/hold (an object) [0.645, 0.144, 0.757, 0.929]", "C. talk to (e.g., self, a person, a group) [0.562, 0.299, 0.565, 0.930] -> put down [0.646, 0.503, 0.807, 1.000] -> carry/hold (an object) [0.471, 0.390, 0.813, 0.964]", "D. walk [0.478, 0.335, 0.635, 0.981] -> carry/hold (an object) [0.478, 0.335, 0.635, 0.981] -> talk to (e.g., self, a person, a group) [0.478, 0.335, 0.635, 0.981]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.478, 0.335, 0.635, 0.981] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1166, "options": [ "A. [0.0, 0.0, 0.388, 1.0]", "B. [0.38, 0.109, 0.772, 0.994]", "C. [0.177, 0.084, 0.591, 0.953]", "D. [0.357, 0.0, 0.464, 0.725]" ], "question": "Where is the person currently performing the open (e.g., a window, a car door) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1311, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1347, "options": [ "A. [0.221, 0.192, 0.396, 0.987]", "B. [0.026, 0.07, 0.937, 0.975]", "C. [0.0, 0.0, 0.242, 1.0]", "D. [0.186, 0.459, 0.26, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1361, "options": [ "A. carry/hold (an object) -> fall down", "B. walk -> carry/hold (an object)", "C. shovel -> carry/hold (an object)", "D. walk -> exit" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1369, "options": [ "A. walk -> carry/hold (an object)", "B. carry/hold (an object) -> enter", "C. kiss (a person) -> walk", "D. drink -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1436, "options": [ "A. stand, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. stand, carry/hold (an object), watch (a person)", "D. stand, talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.0, 0.0, 0.272, 0.998] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1456, "options": [ "A. crouch/kneel [0.676, 0.217, 0.684, 0.683] -> carry/hold (an object) [0.436, 0.191, 0.524, 0.606] -> pull (an object) [0.521, 0.126, 0.508, 0.712]", "B. carry/hold (an object) [0.322, 0.215, 0.427, 0.801] -> walk [0.588, 0.391, 0.651, 0.635] -> take (an object) from (a person) [0.620, 0.400, 0.654, 0.630]", "C. walk [0.487, 0.226, 0.578, 0.643] -> carry/hold (an object) [0.487, 0.226, 0.578, 0.643] -> pull (an object) [0.487, 0.226, 0.578, 0.643]", "D. swim [0.311, 0.143, 0.673, 0.836] -> pull (an object) [0.384, 0.282, 0.766, 0.786] -> walk [0.362, 0.247, 0.705, 0.569]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.487, 0.226, 0.578, 0.643] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1492, "options": [ "A. carry/hold (an object) and lie/sleep", "B. smoke", "C. touch (an object)", "D. hand shake" ], "question": "What action is the person currently in the [0.004, 0.067, 0.993, 0.996] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1494, "options": [ "A. hand shake -> lie/sleep", "B. fishing -> lie/sleep", "C. carry/hold (an object) -> lie/sleep", "D. lift/pick up -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1538, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1571, "options": [ "A. [0.065, 0.04, 0.697, 0.987]", "B. [0.357, 0.395, 0.983, 1.0]", "C. [0.367, 0.097, 0.721, 0.984]", "D. [0.628, 0.0, 0.7, 0.736]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1605, "options": [ "A. [0.301, 0.0, 0.365, 1.0]", "B. [0.655, 0.0, 0.941, 0.99]", "C. [0.33, 0.342, 0.54, 0.988]", "D. [0.456, 0.101, 0.662, 0.988]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1620, "options": [ "A. [0.651, 0.324, 1.0, 1.0]", "B. [0.198, 0.342, 0.867, 1.0]", "C. [0.435, 0.053, 0.961, 0.991]", "D. [0.052, 0.023, 0.904, 1.0]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1633, "options": [ "A. stand, watch (a person)", "B. stand, carry/hold (an object), talk to (e.g., self, a person, a group), watch (a person)", "C. stand, listen to (a person)", "D. stand, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.08, 0.043, 0.427, 0.989] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1644, "options": [ "A. touch (an object) [0.657, 0.334, 0.639, 0.854] -> cut [0.409, 0.247, 0.497, 0.759] -> talk to (e.g., self, a person, a group) [0.584, 0.350, 0.440, 1.000]", "B. touch (an object) [0.371, 0.155, 0.484, 0.997] -> talk to (e.g., self, a person, a group) [0.354, 0.295, 0.567, 1.000] -> ride (e.g., a bike, a car, a horse) [0.521, 0.310, 0.506, 1.000]", "C. carry/hold (an object) [0.423, 0.058, 0.463, 1.000] -> hand clap [0.307, 0.288, 0.787, 1.000] -> talk to (e.g., self, a person, a group) [0.368, 0.309, 0.722, 0.896]", "D. talk to (e.g., self, a person, a group) [0.449, 0.16, 0.636, 0.971] -> carry/hold (an object) [0.464, 0.202, 0.614, 0.955] -> touch (an object) [0.464, 0.202, 0.614, 0.955]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.464, 0.202, 0.614, 0.955] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1731, "options": [ "A. [0.33, 0.342, 0.54, 0.988]", "B. [0.605, 0.035, 1.0, 0.828]", "C. [0.195, 0.0, 1.0, 0.729]", "D. [0.448, 0.162, 0.955, 0.981]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1762, "options": [ "A. swim [0.511, 0.166, 0.765, 0.901] -> walk [0.640, 0.184, 0.518, 1.000] -> talk to (e.g., self, a person, a group) [0.513, 0.000, 0.863, 0.853]", "B. cut [0.600, 0.000, 0.532, 1.000] -> carry/hold (an object) [0.526, 0.079, 0.511, 0.860] -> walk [0.422, 0.000, 0.513, 0.864]", "C. walk [0.503, 0.071, 0.667, 0.994] -> carry/hold (an object) [0.503, 0.071, 0.667, 0.994] -> talk to (e.g., self, a person, a group) [0.503, 0.071, 0.667, 0.994]", "D. carry/hold (an object) [0.399, 0.064, 0.584, 0.796] -> carry/hold (an object) [0.669, 0.060, 0.805, 1.000] -> walk [0.608, 0.136, 0.653, 0.965]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.503, 0.071, 0.667, 0.994] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1786, "options": [ "A. Yes", "B. No" ], "question": "Are there still 6 people in the current frame now?", "sub_answer_type": "Action Persistence" } ]

[ 900, 1786 ]

[ { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 914, "options": [ "A. [0.169, 0.0, 0.741, 0.987]", "B. [0.248, 0.001, 0.982, 0.975]", "C. [0.086, 0.0, 0.72, 0.705]", "D. [0.029, 0.041, 0.543, 1.0]" ], "question": "Where was the person currently performing the smoke in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 929, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 947, "options": [ "A. smoke -> carry/hold (an object)", "B. hug (a person) -> talk to (e.g., self, a person, a group)", "C. brush teeth -> smoke", "D. smoke -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 963, "options": [ "A. talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)", "B. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "C. drive (e.g., a car, a truck) -> carry/hold (an object)", "D. hug (a person) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1049, "options": [ "A. [0.498, 0.033, 0.644, 0.259]", "B. [0.793, 0.0, 0.694, 0.32]", "C. [0.276, 0.369, 0.398, 0.798]", "D. [0.718, 0.0, 0.768, 0.0]" ], "question": "Where is the person currently performing the sail boat located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1104, "options": [ "A. play board game -> play with pets", "B. play board game -> talk to (e.g., self, a person, a group)", "C. play board game -> press", "D. play board game -> put down" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1315, "options": [ "A. [0.0, 0.301, 0.399, 0.982]", "B. [0.0, 0.041, 0.33, 0.942]", "C. [0.165, 0.011, 0.48, 0.782]", "D. [0.0, 0.585, 0.673, 1.0]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1324, "options": [ "A. [0.0, 0.092, 0.302, 1.0]", "B. [0.0, 0.014, 0.271, 0.762]", "C. [0.726, 0.495, 0.845, 0.64]", "D. [0.285, 0.27, 0.414, 1.0]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1376, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.005, 0.059, 0.371, 0.988] location in the current frame performing the open (e.g., a window, a car door)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1382, "options": [ "A. [0.242, 0.227, 0.704, 1.0]", "B. [0.029, 0.108, 0.573, 0.983]", "C. [0.0, 0.339, 0.315, 0.927]", "D. [0.588, 0.219, 0.989, 0.998]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1416, "options": [ "A. hand clap -> walk", "B. touch (an object) -> paint", "C. touch (an object) -> walk", "D. walk -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1487, "options": [ "A. walk, talk to (e.g., self, a person, a group)", "B. stand, carry/hold (an object)", "C. walk, talk to (e.g., self, a person, a group), watch (a person)", "D. walk, carry/hold (an object)" ], "question": "What action is the person at the location [0.127, 0.201, 0.302, 0.835] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1496, "options": [ "A. walk and talk to (e.g., self, a person, a group)", "B. put down", "C. carry/hold (an object)", "D. enter" ], "question": "What action is the person currently in the [0.167, 0.071, 0.615, 0.985] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1528, "options": [ "A. [0.215, 0.073, 0.605, 0.992]", "B. [0.534, 0.321, 0.676, 0.623]", "C. [0.805, 0.213, 0.744, 0.657]", "D. [0.441, 0.057, 0.413, 0.617]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1532, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1578, "options": [ "A. [0.226, 0.003, 0.985, 0.974]", "B. [0.235, 0.24, 0.686, 0.899]", "C. [0.009, 0.007, 0.48, 0.993]", "D. [0.218, 0.0, 0.696, 0.768]" ], "question": "Where was the person currently performing the walk in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1590, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.498, 0.092, 0.702, 0.574] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1597, "options": [ "A. walk, talk to (e.g., self, a person, a group), watch (a person)", "B. walk, talk to (e.g., self, a person, a group)", "C. walk, carry/hold (an object)", "D. stand, drink, watch (a person)" ], "question": "What action is the person at the location [0.012, 0.121, 0.479, 1.0] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1645, "options": [ "A. [0.0, 0.0, 0.64, 1.0]", "B. [0.078, 0.102, 0.935, 0.978]", "C. [0.391, 0.015, 1.0, 1.0]", "D. [0.338, 0.0, 0.678, 0.714]" ], "question": "Where is the person currently performing the drink located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1730, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.0, 0.208, 0.187, 0.993] location in the current frame performing the drink?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1765, "options": [ "A. throw -> touch (an object)", "B. walk -> touch (an object)", "C. fall down -> touch (an object)", "D. touch (an object) -> fishing" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" } ]

[ 900, 1765 ]

[ { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 946, "options": [ "A. drive (e.g., a car, a truck) -> talk to (e.g., self, a person, a group)", "B. talk to (e.g., self, a person, a group) -> cook", "C. hand shake -> talk to (e.g., self, a person, a group)", "D. drive (e.g., a car, a truck) -> play with kids" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 958, "options": [ "A. walk [0.676, 0.494, 0.492, 1.000] -> carry/hold (an object) [0.640, 0.330, 0.827, 0.762] -> lift/pick up [0.674, 0.186, 0.564, 0.875]", "B. walk [0.514, 0.307, 0.637, 0.906] -> carry/hold (an object) [0.514, 0.307, 0.637, 0.906] -> talk to (e.g., self, a person, a group) [0.514, 0.307, 0.637, 0.906]", "C. give/serve (an object) to (a person) [0.639, 0.249, 0.729, 0.724] -> talk to (e.g., self, a person, a group) [0.579, 0.328, 0.826, 1.000] -> carry/hold (an object) [0.333, 0.276, 0.642, 0.809]", "D. carry/hold (an object) [0.642, 0.260, 0.640, 0.856] -> answer phone [0.564, 0.227, 0.737, 0.931] -> walk [0.385, 0.375, 0.661, 0.954]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.514, 0.307, 0.637, 0.906] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 964, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 983, "options": [ "A. [0.15, 0.106, 0.351, 1.0]", "B. [0.247, 0.244, 0.723, 0.634]", "C. [0.308, 0.251, 0.505, 0.905]", "D. [0.738, 0.416, 0.873, 0.966]" ], "question": "Where was the person currently performing the touch (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1026, "options": [ "A. walk, carry/hold (an object)", "B. stand, watch (a person)", "C. stand, carry/hold (an object), listen to (a person)", "D. sit, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.006, 0.1, 0.52, 0.958] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1110, "options": [ "A. bend/bow (at the waist) -> carry/hold (an object) -> touch (an object)", "B. drive (e.g., a car, a truck) -> carry/hold (an object) -> touch (an object)", "C. dress/put on clothing -> carry/hold (an object) -> bend/bow (at the waist)", "D. touch (an object) -> carry/hold (an object) -> chop" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1164, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.304, 0.048, 0.756, 0.905] location in the current frame performing the smoke?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1186, "options": [ "A. get up [0.291, 0.000, 0.819, 0.937] -> talk to (e.g., self, a person, a group) [0.206, 0.288, 0.807, 1.000] -> bend/bow (at the waist) [0.473, 0.285, 0.763, 1.000]", "B. bend/bow (at the waist) [0.379, 0.104, 0.809, 0.958] -> carry/hold (an object) [0.379, 0.104, 0.809, 0.958] -> talk to (e.g., self, a person, a group) [0.379, 0.104, 0.809, 0.958]", "C. smoke [0.341, 0.012, 0.956, 1.000] -> talk to (e.g., self, a person, a group) [0.429, 0.000, 0.710, 1.000] -> carry/hold (an object) [0.214, 0.281, 0.906, 1.000]", "D. read [0.544, 0.000, 0.993, 1.000] -> bend/bow (at the waist) [0.459, 0.000, 0.648, 0.972] -> carry/hold (an object) [0.203, 0.000, 0.754, 0.982]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.379, 0.104, 0.809, 0.958] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1210, "options": [ "A. smoke [0.332, 0.365, 0.475, 0.710] -> walk [0.463, 0.264, 0.567, 0.532] -> sing to (e.g., self, a person, a group) [0.507, 0.402, 0.566, 0.613]", "B. answer phone [0.270, 0.105, 0.667, 0.558] -> carry/hold (an object) [0.579, 0.277, 0.583, 0.623] -> smoke [0.400, 0.138, 0.534, 0.652]", "C. walk [0.555, 0.128, 0.579, 0.692] -> smoke [0.273, 0.077, 0.621, 0.907] -> martial art [0.617, 0.240, 0.483, 0.655]", "D. carry/hold (an object) [0.466, 0.205, 0.621, 0.759] -> smoke [0.466, 0.205, 0.621, 0.759] -> walk [0.399, 0.184, 0.582, 0.678]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.399, 0.184, 0.582, 0.678] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1256, "options": [ "A. lift (a person) -> lie/sleep -> talk to (e.g., self, a person, a group)", "B. lie/sleep -> hand shake -> carry/hold (an object)", "C. carry/hold (an object) -> lie/sleep -> extract", "D. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> lie/sleep" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1257, "options": [ "A. [0.056, 0.096, 0.819, 0.987]", "B. [0.43, 0.0, 0.973, 0.741]", "C. [0.346, 0.062, 0.835, 0.98]", "D. [0.638, 0.052, 1.0, 0.996]" ], "question": "Where is the person currently performing the carry/hold (an object) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1348, "options": [ "A. carry/hold (an object) [0.304, 0.102, 0.775, 0.801] -> cook [0.387, 0.210, 0.857, 0.820] -> talk to (e.g., self, a person, a group) [0.314, 0.181, 0.824, 0.787]", "B. carry/hold (an object) [0.452, 0.02, 0.963, 0.979] -> give/serve (an object) to (a person) [0.452, 0.02, 0.963, 0.979] -> talk to (e.g., self, a person, a group) [0.452, 0.02, 0.963, 0.979]", "C. give/serve (an object) to (a person) [0.542, 0.000, 1.000, 0.852] -> jump/leap [0.482, 0.048, 0.804, 0.938] -> carry/hold (an object) [0.479, 0.203, 1.000, 1.000]", "D. give/serve (an object) to (a person) [0.506, 0.138, 1.000, 0.779] -> hand clap [0.544, 0.000, 1.000, 0.926] -> talk to (e.g., self, a person, a group) [0.566, 0.000, 1.000, 0.900]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.452, 0.02, 0.963, 0.979] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1349, "options": [ "A. Yes", "B. No" ], "question": "Are there still 2 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "D", "answer_type": "Future Prediction", "middle_frame_timestamp": 1352, "options": [ "A. push (an object)", "B. turn (e.g., a screwdriver)", "C. carry/hold (an object) and talk to (e.g., self, a person, a group)", "D. carry/hold (an object) and give/serve (an object) to (a person) and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.411, 0.031, 0.98, 0.989] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1361, "options": [ "A. [0.29, 0.076, 0.666, 0.979]", "B. [0.36, 0.337, 1.0, 1.0]", "C. [0.394, 0.046, 0.989, 0.96]", "D. [0.408, 0.0, 0.782, 0.671]" ], "question": "Where was the person currently performing the drink in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1371, "options": [ "A. [0.027, 0.06, 0.625, 0.971]", "B. [0.0, 0.134, 0.387, 0.734]", "C. [0.43, 0.68, 0.543, 0.983]", "D. [0.0, 0.333, 0.693, 1.0]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1386, "options": [ "A. lie/sleep -> open (e.g., a window, a car door)", "B. lie/sleep -> walk", "C. carry/hold (an object) -> lie/sleep", "D. carry/hold (an object) -> carry/hold (an object)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1454, "options": [ "A. sit, talk to (e.g., self, a person, a group), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. stand, carry/hold (an object), listen to (a person), watch (a person)", "D. sit, stand, carry/hold (an object), talk to (e.g., self, a person, a group)" ], "question": "What action is the person at the location [0.157, 0.137, 0.524, 0.959] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1539, "options": [ "A. exit [0.249, 0.222, 0.835, 0.864] -> talk to (e.g., self, a person, a group) [0.000, 0.000, 0.609, 1.000] -> read [0.000, 0.117, 0.802, 0.899]", "B. carry/hold (an object) [0.074, 0.052, 0.755, 0.976] -> read [0.074, 0.052, 0.755, 0.976] -> talk to (e.g., self, a person, a group) [0.074, 0.052, 0.755, 0.976]", "C. talk to (e.g., self, a person, a group) [0.000, 0.122, 0.642, 1.000] -> dress/put on clothing [0.000, 0.173, 0.780, 0.974] -> carry/hold (an object) [0.045, 0.000, 0.901, 1.000]", "D. talk to (e.g., self, a person, a group) [0.185, 0.214, 0.763, 0.985] -> read [0.000, 0.097, 0.883, 1.000] -> crouch/kneel [0.043, 0.000, 0.720, 0.995]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.074, 0.052, 0.755, 0.976] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1542, "options": [ "A. carry/hold (an object) -> read -> talk to (e.g., self, a person, a group)", "B. carry/hold (an object) -> answer phone -> read", "C. carry/hold (an object) -> talk to (e.g., self, a person, a group) -> kiss (a person)", "D. dress/put on clothing -> carry/hold (an object) -> read" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1572, "options": [ "A. [0.358, 0.459, 0.559, 0.845]", "B. [0.57, 0.054, 0.989, 0.967]", "C. [0.758, 0.0, 1.0, 0.757]", "D. [0.429, 0.0, 0.791, 0.753]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1669, "options": [ "A. 5", "B. 7", "C. 4", "D. 3" ], "question": "How many people were performing the stand in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1738, "options": [ "A. 4", "B. 6", "C. 3", "D. 5" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1755, "options": [ "A. [0.588, 0.081, 0.749, 0.885]", "B. [0.287, 0.506, 1.0, 1.0]", "C. [0.004, 0.3, 0.997, 0.993]", "D. [0.0, 0.41, 0.71, 1.0]" ], "question": "Where is the person currently performing the lie/sleep located in the picture?", "sub_answer_type": "Action Location" } ]

[ 900, 1755 ]

[ { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 917, "options": [ "A. [0.0, 0.124, 0.453, 0.919]", "B. [0.002, 0.332, 0.174, 0.834]", "C. [0.284, 0.329, 0.567, 0.706]", "D. [0.504, 0.167, 0.922, 0.913]" ], "question": "Where is the person currently performing the walk located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 920, "options": [ "A. [0.304, 0.0, 0.303, 0.777]", "B. [0.24, 0.36, 0.598, 0.901]", "C. [0.01, 0.156, 0.352, 0.892]", "D. [0.0, 0.034, 0.219, 0.606]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1068, "options": [ "A. Yes", "B. No" ], "question": "Are there still 1 people in the current frame now?", "sub_answer_type": "Action Persistence" }, { "answer": "A", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1175, "options": [ "A. dance [0.577, 0.266, 0.779, 0.899] -> hug (a person) [0.577, 0.266, 0.779, 0.899] -> sing to (e.g., self, a person, a group) [0.577, 0.266, 0.779, 0.899]", "B. hug (a person) [0.469, 0.222, 0.594, 0.821] -> dance [0.489, 0.138, 0.841, 0.831] -> swim [0.531, 0.318, 0.656, 1.000]", "C. dance [0.765, 0.266, 0.664, 0.853] -> hug (a person) [0.709, 0.374, 0.930, 1.000] -> talk to (e.g., self, a person, a group) [0.569, 0.292, 0.735, 1.000]", "D. walk [0.466, 0.300, 0.635, 0.768] -> hug (a person) [0.564, 0.399, 0.792, 0.742] -> sing to (e.g., self, a person, a group) [0.604, 0.200, 0.632, 0.936]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.577, 0.266, 0.779, 0.899] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1210, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.537, 0.279, 0.769, 0.898] location in the current frame performing the walk?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1211, "options": [ "A. 3", "B. 2", "C. 4", "D. 5" ], "question": "How many people in the current frame are performing the action: dance?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1237, "options": [ "A. swim [0.226, 0.280, 0.471, 0.898] -> crouch/kneel [0.208, 0.312, 0.558, 0.966] -> carry/hold (an object) [0.191, 0.346, 0.195, 0.890]", "B. carry/hold (an object) [0.223, 0.459, 0.361, 0.864] -> crouch/kneel [0.223, 0.459, 0.361, 0.864] -> give/serve (an object) to (a person) [0.223, 0.459, 0.361, 0.864]", "C. hand shake [0.281, 0.471, 0.333, 0.756] -> carry/hold (an object) [0.371, 0.496, 0.194, 0.990] -> crouch/kneel [0.259, 0.572, 0.228, 0.762]", "D. give/serve (an object) to (a person) [0.033, 0.477, 0.474, 0.838] -> carry/hold (an object) [0.156, 0.376, 0.270, 1.000] -> take (an object) from (a person) [0.221, 0.322, 0.197, 0.895]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.223, 0.459, 0.361, 0.864] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1292, "options": [ "A. 4", "B. 2", "C. 5", "D. 3" ], "question": "How many people in the current frame are performing the action: eat?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1332, "options": [ "A. walk [0.581, 0.076, 0.898, 0.767] -> give/serve (an object) to (a person) [0.466, 0.292, 1.000, 0.913] -> carry/hold (an object) [0.718, 0.378, 0.827, 0.805]", "B. bend/bow (at the waist) [0.619, 0.191, 0.883, 0.808] -> carry/hold (an object) [0.619, 0.191, 0.883, 0.808] -> give/serve (an object) to (a person) [0.619, 0.191, 0.883, 0.808]", "C. carry/hold (an object) [0.614, 0.035, 0.858, 0.705] -> hug (a person) [0.559, 0.039, 1.000, 0.623] -> bend/bow (at the waist) [0.699, 0.187, 0.833, 0.810]", "D. bend/bow (at the waist) [0.736, 0.275, 0.798, 0.886] -> carry/hold (an object) [0.714, 0.089, 0.942, 0.953] -> lie/sleep [0.634, 0.114, 1.000, 0.722]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.619, 0.191, 0.883, 0.808] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1333, "options": [ "A. give/serve (an object) to (a person) [0.622, 0.168, 0.738, 0.761] -> bend/bow (at the waist) [0.511, 0.239, 1.000, 0.767] -> crouch/kneel [0.581, 0.198, 1.000, 0.842]", "B. bend/bow (at the waist) [0.619, 0.191, 0.883, 0.808] -> carry/hold (an object) [0.619, 0.191, 0.883, 0.808] -> give/serve (an object) to (a person) [0.619, 0.191, 0.883, 0.808]", "C. swim [0.748, 0.234, 1.000, 0.818] -> bend/bow (at the waist) [0.657, 0.052, 0.778, 0.990] -> give/serve (an object) to (a person) [0.782, 0.216, 1.000, 0.622]", "D. carry/hold (an object) [0.569, 0.335, 0.855, 0.928] -> shovel [0.438, 0.241, 0.690, 0.609] -> give/serve (an object) to (a person) [0.789, 0.340, 1.000, 0.911]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.619, 0.191, 0.883, 0.808] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Future Prediction", "middle_frame_timestamp": 1365, "options": [ "A. clink glass", "B. ride (e.g., a bike, a car, a horse)", "C. crawl", "D. eat" ], "question": "What action is the person currently in the [0.026, 0.12, 0.515, 0.898] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1437, "options": [ "A. 3", "B. 4", "C. 5", "D. 2" ], "question": "How many people in the current frame are performing the action: walk?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1476, "options": [ "A. 4", "B. 3", "C. 8", "D. 5" ], "question": "How many people were performing the listen to (a person) in the scene 60 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1569, "options": [ "A. 4", "B. 3", "C. 2", "D. 5" ], "question": "How many people in the current frame are performing the action: talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Location" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1624, "options": [ "A. 6", "B. 9", "C. 4", "D. 3" ], "question": "How many people were performing the stand in the scene 120 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1654, "options": [ "A. 4", "B. 3", "C. 6", "D. 9" ], "question": "How many people were performing the stand in the scene 180 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1735, "options": [ "A. open (e.g., a window, a car door) [0.255, 0.000, 0.600, 1.000] -> touch (an object) [0.306, 0.206, 0.742, 0.853] -> talk to (e.g., self, a person, a group) [0.310, 0.278, 0.450, 1.000]", "B. talk to (e.g., self, a person, a group) [0.313, 0.123, 0.588, 0.896] -> bend/bow (at the waist) [0.295, 0.139, 0.601, 0.904] -> touch (an object) [0.295, 0.139, 0.601, 0.904]", "C. talk to (e.g., self, a person, a group) [0.459, 0.119, 0.632, 1.000] -> drive (e.g., a car, a truck) [0.272, 0.292, 0.745, 0.882] -> bend/bow (at the waist) [0.360, 0.066, 0.600, 0.868]", "D. drink [0.143, 0.216, 0.407, 1.000] -> talk to (e.g., self, a person, a group) [0.497, 0.135, 0.708, 0.887] -> touch (an object) [0.398, 0.333, 0.506, 0.859]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.295, 0.139, 0.601, 0.904] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1795, "options": [ "A. [0.452, 0.099, 0.799, 0.938]", "B. [0.307, 0.386, 0.258, 0.644]", "C. [0.039, 0.45, 0.282, 0.893]", "D. [0.0, 0.75, 0.173, 0.907]" ], "question": "Where was the person currently performing the drink in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1795 ]

[ { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 937, "options": [ "A. talk to (e.g., self, a person, a group) -> walk -> paint", "B. touch (an object) -> walk -> talk to (e.g., self, a person, a group)", "C. talk to (e.g., self, a person, a group) -> carry/hold (an object) -> walk", "D. talk to (e.g., self, a person, a group) -> walk -> crouch/kneel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 946, "options": [ "A. bend/bow (at the waist), stand, listen to (a person), watch (a person)", "B. stand, listen to (a person), watch (a person)", "C. walk, carry/hold (an object)", "D. stand, talk to (e.g., self, a person, a group), watch (a person)" ], "question": "What action is the person at the location [0.042, 0.246, 0.386, 0.988] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 958, "options": [ "A. lie/sleep [0.011, 0.077, 0.240, 0.971] -> talk to (e.g., self, a person, a group) [0.000, 0.077, 0.438, 0.923] -> bend/bow (at the waist) [0.000, 0.304, 0.425, 1.000]", "B. bend/bow (at the waist) [0.231, 0.274, 0.585, 1.000] -> talk to (e.g., self, a person, a group) [0.022, 0.277, 0.314, 0.948] -> grab (a person) [0.000, 0.367, 0.511, 0.983]", "C. talk to (e.g., self, a person, a group) [0.003, 0.272, 0.419, 0.995] -> bend/bow (at the waist) [0.033, 0.217, 0.424, 0.975] -> grab (a person) [0.033, 0.217, 0.424, 0.975]", "D. grab (a person) [0.120, 0.296, 0.324, 1.000] -> cut [0.000, 0.137, 0.369, 0.874] -> talk to (e.g., self, a person, a group) [0.000, 0.076, 0.437, 1.000]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.033, 0.217, 0.424, 0.975] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1032, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.213, 0.119, 0.406, 0.94] location in the current frame performing the carry/hold (an object)?", "sub_answer_type": "Action Persistence" }, { "answer": "B", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1042, "options": [ "A. smoke [0.291, 0.174, 0.972, 0.835] -> walk [0.256, 0.097, 0.849, 1.000] -> carry/hold (an object) [0.000, 0.000, 0.988, 1.000]", "B. walk [0.13, 0.03, 0.958, 0.99] -> carry/hold (an object) [0.13, 0.03, 0.958, 0.99] -> talk to (e.g., self, a person, a group) [0.13, 0.03, 0.958, 0.99]", "C. carry/hold (an object) [0.000, 0.000, 0.909, 1.000] -> lift (a person) [0.023, 0.032, 0.825, 1.000] -> walk [0.162, 0.153, 0.897, 1.000]", "D. carry/hold (an object) [0.000, 0.221, 1.000, 0.940] -> enter [0.211, 0.062, 1.000, 1.000] -> talk to (e.g., self, a person, a group) [0.000, 0.000, 0.812, 0.943]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.13, 0.03, 0.958, 0.99] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Past Memory", "middle_frame_timestamp": 1110, "options": [ "A. [0.057, 0.087, 0.191, 0.751]", "B. [0.553, 0.315, 0.795, 0.98]", "C. [0.289, 0.379, 0.727, 1.0]", "D. [0.504, 0.069, 0.605, 1.0]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Past Memory", "middle_frame_timestamp": 1127, "options": [ "A. [0.215, 0.048, 0.428, 0.95]", "B. [0.304, 0.326, 0.586, 1.0]", "C. [0.377, 0.026, 0.736, 0.99]", "D. [0.546, 0.0, 0.988, 0.912]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "A", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1170, "options": [ "A. [0.255, 0.047, 0.994, 0.968]", "B. [0.0, 0.018, 0.865, 0.961]", "C. [0.53, 0.054, 0.933, 0.691]", "D. [0.54, 0.218, 1.0, 0.995]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1365, "options": [ "A. carry/hold (an object) [0.673, 0.232, 0.941, 0.845] -> take (an object) from (a person) [0.686, 0.000, 0.927, 0.879] -> walk [0.664, 0.001, 0.741, 0.828]", "B. walk [0.483, 0.113, 0.785, 0.941] -> carry/hold (an object) [0.416, 0.000, 0.831, 1.000] -> ride (e.g., a bike, a car, a horse) [0.643, 0.195, 0.764, 0.929]", "C. talk to (e.g., self, a person, a group) [0.583, 0.133, 1.000, 0.949] -> carry/hold (an object) [0.535, 0.028, 0.811, 1.000] -> kick (a person) [0.648, 0.175, 1.000, 0.887]", "D. walk [0.608, 0.038, 0.887, 0.986] -> carry/hold (an object) [0.608, 0.038, 0.887, 0.986] -> talk to (e.g., self, a person, a group) [0.608, 0.038, 0.887, 0.986]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.608, 0.038, 0.887, 0.986] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "B", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1377, "options": [ "A. write -> carry/hold (an object)", "B. talk to (e.g., self, a person, a group) -> carry/hold (an object)", "C. carry/hold (an object) -> martial art", "D. talk to (e.g., self, a person, a group) -> ride (e.g., a bike, a car, a horse)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1390, "options": [ "A. bend/bow (at the waist) [0.295, 0.041, 0.648, 1.000] -> clink glass [0.112, 0.202, 0.546, 0.804] -> carry/hold (an object) [0.285, 0.000, 0.688, 0.894]", "B. carry/hold (an object) [0.135, 0.000, 0.683, 0.913] -> point to (an object) [0.113, 0.000, 0.553, 0.796] -> bend/bow (at the waist) [0.200, 0.155, 0.697, 1.000]", "C. bend/bow (at the waist) [0.053, 0.039, 0.530, 0.764] -> carry/hold (an object) [0.212, 0.161, 0.538, 1.000] -> walk [0.092, 0.000, 0.392, 1.000]", "D. bend/bow (at the waist) [0.186, 0.061, 0.508, 0.96] -> carry/hold (an object) [0.186, 0.061, 0.508, 0.96] -> talk to (e.g., self, a person, a group) [0.186, 0.061, 0.508, 0.96]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.186, 0.061, 0.508, 0.96] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "D", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1430, "options": [ "A. [0.0, 0.0, 0.771, 0.68]", "B. [0.322, 0.021, 0.978, 1.0]", "C. [0.227, 0.412, 0.93, 1.0]", "D. [0.027, 0.122, 0.89, 0.968]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1445, "options": [ "A. [0.725, 0.108, 0.874, 0.856]", "B. [0.227, 0.0, 0.931, 0.804]", "C. [0.147, 0.194, 0.973, 1.0]", "D. [0.011, 0.033, 0.831, 0.991]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1445, "options": [ "A. read -> talk to (e.g., self, a person, a group)", "B. open (e.g., a window, a car door) -> crouch/kneel", "C. talk to (e.g., self, a person, a group) -> point to (an object)", "D. talk to (e.g., self, a person, a group) -> crouch/kneel" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "A", "answer_type": "Future Prediction", "middle_frame_timestamp": 1498, "options": [ "A. crouch/kneel and talk to (e.g., self, a person, a group)", "B. chop", "C. touch (an object)", "D. carry/hold (an object) and talk to (e.g., self, a person, a group)" ], "question": "What action is the person currently in the [0.141, 0.616, 0.706, 1.0] location likely to do next?", "sub_answer_type": "Action Anticipation" }, { "answer": "B", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1514, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.239, 0.048, 1.0, 1.0] location in the current frame performing the bend/bow (at the waist)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1517, "options": [ "A. cook -> talk to (e.g., self, a person, a group) -> bend/bow (at the waist)", "B. bend/bow (at the waist) -> talk to (e.g., self, a person, a group) -> walk", "C. crouch/kneel -> talk to (e.g., self, a person, a group) -> bend/bow (at the waist)", "D. hit (an object) -> bend/bow (at the waist) -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "D", "answer_type": "Temporal Perception", "middle_frame_timestamp": 1555, "options": [ "A. lie/sleep -> run/jog", "B. talk to (e.g., self, a person, a group) -> martial art", "C. press -> run/jog", "D. run/jog -> talk to (e.g., self, a person, a group)" ], "question": "What is the sequence of actions the person in the scene has performed recently?", "sub_answer_type": "Action Sequence" }, { "answer": "B", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1572, "options": [ "A. stand, listen to (a person), watch (a person)", "B. walk, carry/hold (an object)", "C. stand, talk to (e.g., self, a person, a group), watch (a person)", "D. sit, listen to (a person), watch (a person)" ], "question": "What action is the person at the location [0.509, 0.188, 0.78, 0.923] in the image performing?", "sub_answer_type": "Action Location" }, { "answer": "A", "answer_type": "Temporal Hallucination Verification", "middle_frame_timestamp": 1629, "options": [ "A. Yes", "B. No" ], "question": "Is the person currently in the [0.149, 0.125, 0.929, 1.0] location in the current frame performing the talk to (e.g., self, a person, a group)?", "sub_answer_type": "Action Persistence" }, { "answer": "C", "answer_type": "Spatio Perception", "middle_frame_timestamp": 1633, "options": [ "A. [0.022, 0.052, 0.774, 0.974]", "B. [0.0, 0.294, 0.937, 1.0]", "C. [0.116, 0.06, 1.0, 0.992]", "D. [0.152, 0.0, 0.874, 0.703]" ], "question": "Where is the person currently performing the talk to (e.g., self, a person, a group) located in the picture?", "sub_answer_type": "Action Location" }, { "answer": "D", "answer_type": "Past Memory", "middle_frame_timestamp": 1644, "options": [ "A. [0.064, 0.426, 0.493, 1.0]", "B. [0.069, 0.081, 0.22, 0.747]", "C. [0.596, 0.563, 0.785, 0.909]", "D. [0.318, 0.344, 0.576, 0.889]" ], "question": "Where was the person currently performing the talk to (e.g., self, a person, a group) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" }, { "answer": "C", "answer_type": "Spatio Temporal Perception", "middle_frame_timestamp": 1657, "options": [ "A. dance [0.351, 0.101, 0.519, 0.867] -> carry/hold (an object) [0.511, 0.279, 0.651, 0.940] -> shoot [0.493, 0.341, 0.749, 0.788]", "B. write [0.284, 0.123, 0.877, 0.856] -> carry/hold (an object) [0.508, 0.131, 0.735, 0.923] -> sing to (e.g., self, a person, a group) [0.405, 0.348, 0.854, 0.878]", "C. carry/hold (an object) [0.33, 0.163, 0.712, 0.983] -> dance [0.33, 0.163, 0.712, 0.983] -> sing to (e.g., self, a person, a group) [0.33, 0.163, 0.712, 0.983]", "D. carry/hold (an object) [0.269, 0.003, 0.680, 0.824] -> close (e.g., a door, a box) [0.433, 0.195, 0.527, 1.000] -> sing to (e.g., self, a person, a group) [0.365, 0.131, 0.622, 0.784]" ], "question": "What is the sequence of actions and the corresponding movement trajectory of the person currently in the [0.33, 0.163, 0.712, 0.983] location?", "sub_answer_type": "Action Trajectory" }, { "answer": "A", "answer_type": "Past Memory", "middle_frame_timestamp": 1762, "options": [ "A. [0.073, 0.077, 0.229, 0.748]", "B. [0.259, 0.051, 0.546, 0.972]", "C. [0.0, 0.0, 0.447, 0.531]", "D. [0.0, 0.272, 0.163, 1.0]" ], "question": "Where was the person currently performing the carry/hold (an object) in the scene 8 seconds ago?", "sub_answer_type": "Action Retrieval" } ]

[ 900, 1762 ]