MediaPipe-Face-Detection: Optimized for Mobile Deployment
Detect faces and locate facial features in real-time video and image streams
Designed for sub-millisecond processing, this model predicts bounding boxes and pose skeletons (left eye, right eye, nose tip, mouth, left eye tragion, and right eye tragion) of faces in an image.
This model is an implementation of MediaPipe-Face-Detection found here.
This repository provides scripts to run MediaPipe-Face-Detection on Qualcomm® devices. More details on model performance across various devices, can be found here.
Model Details
- Model Type: Object detection
- Model Stats:
- Input resolution: 256x256
- Number of parameters (MediaPipeFaceDetector): 135K
- Model size (MediaPipeFaceDetector): 565 KB
- Number of parameters (MediaPipeFaceLandmarkDetector): 603K
- Model size (MediaPipeFaceLandmarkDetector): 2.34 MB
- Number of output classes: 6
Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model |
---|---|---|---|---|---|---|---|---|
MediaPipeFaceDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.554 ms | 0 - 7 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.543 ms | 1 - 4 MB | FP16 | NPU | MediaPipe-Face-Detection.so |
MediaPipeFaceDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 0.862 ms | 1 - 7 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.398 ms | 0 - 27 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.38 ms | 1 - 19 MB | FP16 | NPU | MediaPipe-Face-Detection.so |
MediaPipeFaceDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.597 ms | 0 - 23 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.35 ms | 0 - 25 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.469 ms | 1 - 26 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.625 ms | 1 - 28 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceDetector | SA7255P ADP | SA7255P | TFLITE | 18.735 ms | 0 - 14 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | SA7255P ADP | SA7255P | QNN | 18.67 ms | 1 - 9 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.556 ms | 0 - 6 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.544 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | SA8295P ADP | SA8295P | TFLITE | 1.141 ms | 0 - 18 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | SA8295P ADP | SA8295P | QNN | 1.082 ms | 0 - 14 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.557 ms | 0 - 6 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.536 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | SA8775P ADP | SA8775P | TFLITE | 1.28 ms | 0 - 15 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | SA8775P ADP | SA8775P | QNN | 1.207 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 18.735 ms | 0 - 14 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 18.67 ms | 1 - 9 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.556 ms | 0 - 6 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.534 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 1.28 ms | 0 - 15 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 1.207 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.744 ms | 0 - 19 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceDetector | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.747 ms | 1 - 24 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.661 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.849 ms | 2 - 2 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.2 ms | 0 - 8 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.221 ms | 0 - 3 MB | FP16 | NPU | MediaPipe-Face-Detection.so |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 0.375 ms | 0 - 9 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.149 ms | 0 - 27 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.163 ms | 0 - 18 MB | FP16 | NPU | MediaPipe-Face-Detection.so |
MediaPipeFaceLandmarkDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.277 ms | 0 - 26 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceLandmarkDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.127 ms | 0 - 20 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.141 ms | 0 - 19 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.295 ms | 0 - 20 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
MediaPipeFaceLandmarkDetector | SA7255P ADP | SA7255P | TFLITE | 3.562 ms | 0 - 12 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | SA7255P ADP | SA7255P | QNN | 3.59 ms | 0 - 8 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.199 ms | 0 - 8 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.221 ms | 0 - 3 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | SA8295P ADP | SA8295P | TFLITE | 0.569 ms | 0 - 18 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | SA8295P ADP | SA8295P | QNN | 0.592 ms | 0 - 14 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.197 ms | 0 - 8 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.219 ms | 0 - 2 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | SA8775P ADP | SA8775P | TFLITE | 0.51 ms | 0 - 13 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | SA8775P ADP | SA8775P | QNN | 0.519 ms | 0 - 11 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 3.562 ms | 0 - 12 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 3.59 ms | 0 - 8 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.2 ms | 0 - 9 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.218 ms | 0 - 3 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 0.51 ms | 0 - 13 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 0.519 ms | 0 - 11 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.275 ms | 0 - 18 MB | FP16 | NPU | MediaPipe-Face-Detection.tflite |
MediaPipeFaceLandmarkDetector | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.297 ms | 0 - 23 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.328 ms | 0 - 0 MB | FP16 | NPU | Use Export Script |
MediaPipeFaceLandmarkDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.375 ms | 2 - 2 MB | FP16 | NPU | MediaPipe-Face-Detection.onnx |
Installation
Install the package via pip:
pip install qai-hub-models
Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
Sign-in to Qualcomm® AI Hub with your
Qualcomm® ID. Once signed in navigate to Account -> Settings -> API Token
.
With this API token, you can configure your client to run models on the cloud hosted devices.
qai-hub configure --api_token API_TOKEN
Navigate to docs for more information.
Demo off target
The package contains a simple end-to-end demo that downloads pre-trained weights and runs this model on a sample input.
python -m qai_hub_models.models.mediapipe_face.demo
The above demo runs a reference implementation of pre-processing, model inference, and post processing.
NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).
%run -m qai_hub_models.models.mediapipe_face.demo
Run model on a cloud-hosted device
In addition to the demo, you can also run the model on a cloud-hosted Qualcomm® device. This script does the following:
- Performance check on-device on a cloud-hosted device
- Downloads compiled assets that can be deployed on-device for Android.
- Accuracy check between PyTorch and on-device outputs.
python -m qai_hub_models.models.mediapipe_face.export
Profiling Results
------------------------------------------------------------
MediaPipeFaceDetector
Device : Samsung Galaxy S23 (13)
Runtime : TFLITE
Estimated inference time (ms) : 0.6
Estimated peak memory usage (MB): [0, 7]
Total # Ops : 111
Compute Unit(s) : NPU (111 ops)
------------------------------------------------------------
MediaPipeFaceLandmarkDetector
Device : Samsung Galaxy S23 (13)
Runtime : TFLITE
Estimated inference time (ms) : 0.2
Estimated peak memory usage (MB): [0, 8]
Total # Ops : 100
Compute Unit(s) : NPU (100 ops)
How does this work?
This export script leverages Qualcomm® AI Hub to optimize, validate, and deploy this model on-device. Lets go through each step below in detail:
Step 1: Compile model for on-device deployment
To compile a PyTorch model for on-device deployment, we first trace the model
in memory using the jit.trace
and then call the submit_compile_job
API.
import torch
import qai_hub as hub
from qai_hub_models.models.mediapipe_face import Model
# Load the model
model = Model.from_pretrained()
face_detector_model = model.face_detector
face_landmark_detector_model = model.face_landmark_detector
# Device
device = hub.Device("Samsung Galaxy S23")
# Trace model
face_detector_input_shape = face_detector_model.get_input_spec()
face_detector_sample_inputs = face_detector_model.sample_inputs()
traced_face_detector_model = torch.jit.trace(face_detector_model, [torch.tensor(data[0]) for _, data in face_detector_sample_inputs.items()])
# Compile model on a specific device
face_detector_compile_job = hub.submit_compile_job(
model=traced_face_detector_model ,
device=device,
input_specs=face_detector_model.get_input_spec(),
)
# Get target model to run on-device
face_detector_target_model = face_detector_compile_job.get_target_model()
# Trace model
face_landmark_detector_input_shape = face_landmark_detector_model.get_input_spec()
face_landmark_detector_sample_inputs = face_landmark_detector_model.sample_inputs()
traced_face_landmark_detector_model = torch.jit.trace(face_landmark_detector_model, [torch.tensor(data[0]) for _, data in face_landmark_detector_sample_inputs.items()])
# Compile model on a specific device
face_landmark_detector_compile_job = hub.submit_compile_job(
model=traced_face_landmark_detector_model ,
device=device,
input_specs=face_landmark_detector_model.get_input_spec(),
)
# Get target model to run on-device
face_landmark_detector_target_model = face_landmark_detector_compile_job.get_target_model()
Step 2: Performance profiling on cloud-hosted device
After compiling models from step 1. Models can be profiled model on-device using the
target_model
. Note that this scripts runs the model on a device automatically
provisioned in the cloud. Once the job is submitted, you can navigate to a
provided job URL to view a variety of on-device performance metrics.
face_detector_profile_job = hub.submit_profile_job(
model=face_detector_target_model,
device=device,
)
face_landmark_detector_profile_job = hub.submit_profile_job(
model=face_landmark_detector_target_model,
device=device,
)
Step 3: Verify on-device accuracy
To verify the accuracy of the model on-device, you can run on-device inference on sample input data on the same cloud hosted device.
face_detector_input_data = face_detector_model.sample_inputs()
face_detector_inference_job = hub.submit_inference_job(
model=face_detector_target_model,
device=device,
inputs=face_detector_input_data,
)
face_detector_inference_job.download_output_data()
face_landmark_detector_input_data = face_landmark_detector_model.sample_inputs()
face_landmark_detector_inference_job = hub.submit_inference_job(
model=face_landmark_detector_target_model,
device=device,
inputs=face_landmark_detector_input_data,
)
face_landmark_detector_inference_job.download_output_data()
With the output of the model, you can compute like PSNR, relative errors or spot check the output with expected output.
Note: This on-device profiling and inference requires access to Qualcomm® AI Hub. Sign up for access.
Deploying compiled model to Android
The models can be deployed using multiple runtimes:
TensorFlow Lite (
.tflite
export): This tutorial provides a guide to deploy the .tflite model in an Android application.QNN (
.so
export ): This sample app provides instructions on how to use the.so
shared library in an Android application.
View on Qualcomm® AI Hub
Get more details on MediaPipe-Face-Detection's performance across various devices here. Explore all available models on Qualcomm® AI Hub
License
- The license for the original implementation of MediaPipe-Face-Detection can be found here.
- The license for the compiled assets for on-device deployment can be found here
References
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.