Skip to content

Latest commit

Β 

History

History
58 lines (39 loc) Β· 2.15 KB

README.md

File metadata and controls

58 lines (39 loc) Β· 2.15 KB

🎭 Painted Skin 🎭

Welcome to the 🎭 Painted Skin 🎭! Get ready to have some fun with faces. πŸ˜„

License

🐾 Preview

image-20231109170433797

πŸͺ„ How to use

  1. Clone the Repository:

    git clone https://github.com/BlackThompson/Painted-Skin.git
  2. Create a New Environment and Install Dependencies:

    • Note: Python version should be >=3.8
    • Note: The versions of torch, torchvision, and torchaudio should align with your CUDA version.
    conda create --name painted_skin python=3.8
    conda activate painted_skin
    pip install -r requirements.txt 
  3. Run gradio_UI.py:

    gradio gradio_UI.py
  4. Click the Local URL

🦾 How it Works

This tool allows you to perform a face swap. Simply follow these steps:

  1. Upload Source:
    • Choose between src_picture and src_video for the face you want to swap.
    • Note: You can only upload one source at a time, and the tool processes one task at a time.
  2. Upload Target:
    • Upload target_picture, and our model will swap the face onto the source image or video.
  3. Submit and Wait:
    • Click the submit button and patiently wait for the process to complete.
    • Keep in mind that face swapping in videos might take some time, so be patient! πŸ˜…

❗ Important Note

  • If you are able to load src_picture but encounter issues with loading src_video, please ensure that you have downloaded FFmpeg and correctly added its bin directory to your system path.
  • Ensure that target_picture is uploaded for the face swap to work effectively.

Enjoy swapping faces and have a good laugh! πŸ˜†

πŸ’Œ Acknowledgements

This repository borrows heavily from facefusion-damo and face-change. Thanks to the authors for sharing their code and models.