Add model card for ReViSE

README.md

@@ -0,0 +1,114 @@
+---
+license: mit
+pipeline_tag: video-to-video
+library_name: transformers
+---
+
+# ReViSE: Towards Reason-Informed Video Editing in Unified Models with Self-Reflective Learning
+
+This repository contains the official implementation of the paper [ReViSE: Towards Reason-Informed Video Editing in Unified Models with Self-Reflective Learning](https://huggingface.co/papers/2512.09924).
+
+ReViSE introduces the Reason-Informed Video Editing (RVE) task, which requires reasoning about physical plausibility and causal dynamics during editing. It proposes a Self-Reflective Reasoning (SRF) framework that unifies generation and evaluation within a single architecture, utilizing an internal VLM for intrinsic feedback. This model significantly enhances editing accuracy and visual fidelity in reason-informed video editing.
+
+**GitHub Repository:** [https://github.com/Liuxinyv/ReViSE](https://github.com/Liuxinyv/ReViSE)
+
+<div align="center">
+  <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/data.png" alt="RVE-Bench abstract" />
+</div>
+
+## Demos
+### Reason-informed video editing
+<div align="center">
+<table style="border-collapse: collapse; width: 60%; font-size: 14px;">
+  
+  <tr>
+    <td colspan="2" style="text-align:center; padding:6px;">
+      <strong>What if the the dog ran into the depth of a forest?</strong>
+    </td>
+  </tr>
+  <tr>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_001.gif" style="width:250px; height:auto;" />
+    </td>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_002.gif" style="width:250px; height:auto;" />
+    </td>
+  </tr>
+
+  <tr>
+    <td colspan="2" style="text-align:center; padding:6px;">
+      <strong>What if the girl’s fragrance gently attracted a delicate butterfly, fluttering toward her?</strong>
+    </td>
+  </tr>
+  <tr>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_003.gif" style="width:250px; height:auto;" />
+    </td>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_004.gif" style="width:250px; height:auto;" />
+    </td>
+  </tr>
+
+  <tr>
+    <td colspan="2" style="text-align:center; padding:6px;">
+      <strong>What if the scene transitioned from a magical night to a dawn, causing the northern lights to fade away?</strong>
+    </td>
+  </tr>
+  <tr>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_005.gif" style="width:250px; height:auto;" />
+    </td>
+    <td style="text-align:center; padding:4px;">
+      <img src="https://github.com/Liuxinyv/ReViSE/raw/main/assets/demo_006.gif" style="width:250px; height:auto;" />
+    </td>
+  </tr>
+
+</table>
+</div>
+
+## Quick Start (Inference)
+To get started with ReViSE inference, follow these steps:
+
+1. Create conda environment
+```bash
+conda create -n revise python=3.10
+conda activate revise
+
+pip install -r pip_requirements.txt
+```
+2. Set up environment variables for CUDA
+```bash
+# For CUDA (adjust path as needed)
+export CUDA_HOME="/usr/local/cuda"
+export PATH="${CUDA_HOME}/bin:${PATH}"
+export LD_LIBRARY_PATH="${CUDA_HOME}/lib64:${LD_LIBRARY_PATH}"
+
+# Add to inference or training script
+export PYTHONPATH="${PWD}:${PWD}/nets/third_party:${PYTHONPATH}"
+```
+3. Downloading checkpoints
+Download our pretrained model checkpoint [here](https://huggingface.co/sophiaa/revise/tree/main/revise_ckpt).
+
+### Inference
+```bash
+# Run inference with sample data
+bash tools/inference/inference.sh
+```
+
+## Acknowledgement
+We would like to thank [Omni-Video](https://github.com/SAIS-FUXI/Omni-Video), [VILA](https://github.com/NVlabs/VILA) and [Wan2.1](https://github.com/Wan-Video/Wan2.1) for their excellent work.
+
+## Citation
+If you find this project useful, please consider citing:
+
+```bibtex
+@misc{liu2025revisereasoninformedvideoediting,
+      title={ReViSE: Towards Reason-Informed Video Editing in Unified Models with Self-Reflective Learning}, 
+      author={Xinyu Liu and Hangjie Yuan and Yujie Wei and Jiazheng Xing and Yujin Han and Jiahao Pan and Yanbiao Ma and Chi-Min Chan and Kang Zhao and Shiwei Zhang and Wenhan Luo and Yike Guo},
+      year={2025},
+      eprint={2512.09924},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2512.09924}, 
+}
+```