Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,89 @@
+---
+license: mit
+task_categories:
+  - visual-question-answering
+language:
+  - en
+tags:
+  - visual-reasoning
+  - VQA
+  - synthetic
+  - domain-robustness
+  - CLEVR
+pretty_name: Super-CLEVR
+size_categories:
+  - 100K<n<1M
+---
+
+# Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning
+
+**[CVPR 2023 Highlight (top 2.5%)]**
+
+Paper: [Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning](https://arxiv.org/abs/2212.00259)
+
+**Authors:** Zhuowan Li, Xingrui Wang, Elias Stengel-Eskin, Adam Kortylewski, Wufei Ma, Benjamin Van Durme, Alan Yuille
+
+## Dataset Description
+
+Super-CLEVR is a synthetic dataset designed to systematically study the **domain robustness** of visual reasoning models across four key factors:
+
+- **Visual complexity** — varying levels of scene and object complexity
+- **Question redundancy** — controlling redundant information in questions
+- **Concept distribution** — shifts in the distribution of visual concepts
+- **Concept compositionality** — novel compositions of known concepts
+
+## Dataset Structure
+
+Super-CLEVR contains **30,000 images** of vehicles (sourced from [UDA-Part](https://github.com/TACJu/UDA-Part)) randomly placed in 3D-rendered scenes, with **10 question-answer pairs per image** (300k QA pairs total). Vehicles include part-level annotations, enabling questions about distinct part attributes.
+
+### Splits
+
+| Split      | Images       |
+|------------|-------------|
+| Train      | 20,000       |
+| Validation | 5,000        |
+| Test       | 5,000        |
+
+### Files
+
+| File | Description |
+|------|-------------|
+| `images.zip` | 30k rendered scene images |
+| `superCLEVR_scenes.json` | Scene annotations (objects, parts, spatial relations) |
+| `superCLEVR_questions_30k.json` | Standard question-answer pairs |
+| `superCLEVR_questions_30k_NoRedundant.json` | Questions with redundancy removed |
+| `superCLEVR_questions_30k_AllRedundant.json` | Questions with maximum redundancy |
+
+## Usage
+
+```python
+from huggingface_hub import hf_hub_download
+
+# Download a specific file
+path = hf_hub_download(
+    repo_id="RyanWW/Super-CLEVR",
+    filename="superCLEVR_questions_30k.json",
+    repo_type="dataset",
+)
+```
+
+## Citation
+
+```bibtex
+@inproceedings{li2023super,
+  title={Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning},
+  author={Li, Zhuowan and Wang, Xingrui and Stengel-Eskin, Elias and Kortylewski, Adam and Ma, Wufei and Van Durme, Benjamin and Yuille, Alan L},
+  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
+  pages={14963--14973},
+  year={2023}
+}
+```
+
+## Links
+
+- **Code:** [github.com/Lizw14/Super-CLEVR](https://github.com/Lizw14/Super-CLEVR)
+- **Paper:** [arxiv.org/abs/2212.00259](https://arxiv.org/abs/2212.00259)
+
+## License
+
+This dataset is released under the [MIT License](https://opensource.org/licenses/MIT).