ROSE / README.md
weiyuchoumou526's picture
Update README.md
954b185 verified

A newer version of the Gradio SDK is available: 5.49.1

Upgrade
metadata
title: ROSE
emoji: πŸš€
colorFrom: blue
colorTo: pink
sdk: gradio
sdk_version: 5.35.0
short_description: 'ROSE: Remove Objects with Side Effects in Videos'
app_file: app.py
pinned: false

Get Started

  1. Install ROSE Dependencies

    You can follow the Dependencies and Installation.

  2. Install Demo Dependencies

cd hugging_face/

# install python dependencies 
pip3 install -r requirements.txt

# Run the demo
python app.py

Usage Guidance

  • Step 1: Upload your video and click the Get video info button.

  • Step 2:

    1. [Optional] Specify the tracking period for the currently added mask by dragging the Track start frame or Track end frame.
    2. Click the image on the left to select the mask area.
      • Click Add mask if you are satisfied with the mask, or
      • [Optional] Click Clear clicks if you want to reselect the mask area, or
      • [Optional] Click Remove mask to remove all masks.
    3. [Optional] Go back to step 2.1 to add another mask.
  • Step 3:

    1. Click the Tracking button to track the masks for the whole video.
    2. Then click Inpainting to get the inpainting results.

You can always refer to the Highlighted Text box on the page for guidance on the next step!

Citation

If you find our repo useful for your research, please consider citing our paper:

@article{miao2025rose,
      title={ROSE: Remove Objects with Side Effects in Videos}, 
      author={Miao, Chenxuan and Feng, Yutong and Zeng, Jianshu and Gao, Zixiang and Liu, Hantang and Yan, Yunfeng and Qi, Donglian and Chen, Xi and Wang, Bin and Zhao, Hengshuang},
      journal={arXiv preprint arXiv:2508.18633},
      year={2025}
}

Acknowledgements

The project harnesses the capabilities from Track Anything, Segment Anything and Cutie. Also the gradio demo page is based on ProPainter's huggingface demo page. Thanks for their awesome works.