Update dataset card with paper link, task category, and sample usage

README.md

@@ -1,20 +1,27 @@
 ---
+language:
+- en
 license: apache-2.0
+size_categories:
+- 1K<n<10K
 task_categories:
-  - video-classification
-language:
-  - en
+- video-text-to-text
 tags:
-  - video-editing
-  - benchmark
-  - reward-model
-size_categories:
-  - n<1K
+- video-editing
+- benchmark
+- reward-model
 ---
 
-# VEFX-Bench: Video Editing Quality Benchmark
+# VEFX-Bench: A Holistic Benchmark for Generic Video Editing and Visual Effects
 
-VEFX-Bench is a benchmark dataset for evaluating video editing model quality. It contains **300 original source videos** with editing instructions across **9 task categories**.
+[Paper](https://huggingface.co/papers/2604.16272) | [GitHub](https://github.com/Visko-Platform/VEFX-Bench) | [Project Page](https://xiangbogaobarry.github.io/VEFX-Bench/) | [Leaderboard](https://vefx-leaderboard.com/)
+
+**VEFX-Bench** is a comprehensive benchmark for evaluating text-driven video editing and visual effects. It is derived from the **VEFX-Dataset**, a human-annotated collection of 5,049 video editing examples across 9 major categories and 32 subcategories. 
+
+The benchmark provides a curated set of 300 video-prompt pairs for standardized comparison, evaluated across three decoupled dimensions:
+- **Instructional Following (IF)**: How well the edit follows the instruction.
+- **Render Quality (RQ)**: Visual clarity and temporal consistency.
+- **Edit Exclusivity (EE)**: Preservation of unrelated content (edit locality).
 
 ## Dataset Structure
 
@@ -22,33 +29,44 @@ VEFX-Bench is a benchmark dataset for evaluating video editing model quality. It
 - `prompts.json`: Editing instructions indexed by video number
 - `benchmark_meta.json`: Metadata including category labels and filename mapping
 
-## Usage
+## Sample Usage
+
+You can use the associated [VEFX-Reward](https://huggingface.co/xiangbog/VEFX-Reward-4B) model to score your video edits using the following code:
 
-1. Download all 300 videos and `prompts.json`
-2. Apply your video editing model following each prompt
-3. Save edited videos as `0000.mp4` through `0299.mp4` (matching index)
-4. Submit the zip to the [VEFX-Leaderboard](https://github.com/YOUR_ORG/VEFX-Leaderboard)
+```python
+from vefx_reward import VEFXReward
 
-## Evaluation Dimensions
+model = VEFXReward("xiangbog/VEFX-Reward-4B", device="cuda")
 
-- **IF (Instructional Following)**: How well the edit follows the instruction (1-4)
-- **RQ (Render Quality)**: Visual quality and temporal consistency (1-4)
-- **EE (Edit Exclusivity)**: Preservation of unrelated content (1-4)
+scores = model.score(
+    original_video="path/to/original_video.mp4",
+    edited_video="path/to/edited_video.mp4",
+    instruction="Remove the woman with the grey backpack walking on the right side of the frame.",
+)
+print(scores)
+# Example Output: {'IF': 2.34, 'RQ': 1.93, 'EE': 1.82, 'Overall': 6.09}
+```
 
 ## Task Categories
 
-| Category | Count |
-|----------|-------|
-| Style Transfer | varies |
-| Object Manipulation | varies |
-| Background Change | varies |
-| Color/Lighting | varies |
-| Motion/Animation | varies |
-| Text/Overlay | varies |
-| Composition | varies |
-| Removal/Inpainting | varies |
-| Complex/Multi-step | varies |
+The benchmark covers 9 major editing categories:
+- Style Transfer
+- Object Manipulation
+- Background Change
+- Color/Lighting
+- Motion/Animation
+- Text/Overlay
+- Composition
+- Removal/Inpainting
+- Complex/Multi-step
 
 ## Citation
 
-Coming soon.
+```bibtex
+@article{gao2025vefxbench,
+  title={VEFX-Bench: Benchmarking Generic Video Editing and Visual Effects},
+  author={Xiangbo Gao and Sicong Jiang and Bangya Liu and Xinghao Chen and Minglai Yang and Siyuan Yang and Mingyang Wu and Jiongze Yu and Qi Zheng and Haozhi Wang and Jiayi Zhang and Jared Yang and Jie Yang and Zihan Wang and Qing Yin and Zhengzhong Tu},
+  journal={arXiv preprint arXiv:2604.16272},
+  year={2026}
+}
+```