Real-time facial animation, powered by First Order Motion Model (FOMM) and FAN-face

This repository is a heavily modified fork of the First Order Motion Model for Image Animation and also incorporates components from FAN-face

Note:

• GPU is needed, and we only test this model using RTX 3060 (6 GB, Laptop).
• This project is only for windows system.

This is a demo of our project. The left side shows the camera feed, and the right side shows the model's real-time output.

git clone https://github.com/20160161simon/realtime-FOMM.git
cd realtime-fomm-stylegan
pip install -r requirements.txt

• Google Drive
• Tsinghua Cloud

Then, place the checkpoint file in the checkpoints folder. There's no need to extract it—keeping the .tar file as-is is fine.

python camera.py

If you want to train your own model, please follow the instructions provided in the First Order Motion Model for Image Animation.

(!!!) Please note that the original dataset downloader is no longer functional. We have fixed the issue and included the updated version in the dataset_downloader folder. For detailed usage instructions, refer to the README.md file inside that folder.