Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,30 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# Visual Haystacks Dataset Card
|
| 6 |
+
|
| 7 |
+
## Dataset details
|
| 8 |
+
Dataset type: Visual Haystacks (VHs) is a benchmark dataset specifically designed to evaluate Large Multimodal Model's (LMM's) capability to handle long-context visual information. It can also be viewed as the first visual-centric Needle-In-A-Haystack (NIAH) benchmark dataset. Please also download COCO-2017's training set validation set.
|
| 9 |
+
|
| 10 |
+
```
|
| 11 |
+
dataset/
|
| 12 |
+
└── coco
|
| 13 |
+
├── annotations
|
| 14 |
+
├── test2017
|
| 15 |
+
├── train2017
|
| 16 |
+
└── val2017
|
| 17 |
+
```
|
| 18 |
+
|
| 19 |
+
## Dataset date: VHs was collected in April 2024, directly derived from COCO's image and object annotations.
|
| 20 |
+
|
| 21 |
+
## Paper or resources for more information: [TODO]
|
| 22 |
+
|
| 23 |
+
## License: [TODO]
|
| 24 |
+
|
| 25 |
+
Where to send questions or comments about the model: https://github.com/visual-haystacks/[TODO]/issues
|
| 26 |
+
|
| 27 |
+
## Intended use
|
| 28 |
+
Primary intended uses: The primary use of VHs is research on large multimodal models and chatbots.
|
| 29 |
+
|
| 30 |
+
Primary intended users: The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.
|