MyNiuuu/MOFA-Video
An image-to-video diffusion model that animates static images into videos using learned motion field adaptions.
Not currently ranked — collecting fresh signals.
star history
MOFA-Video is a research project from the University of Tokyo and Tencent AI Lab that enables controllable image animation using a frozen image-to-video diffusion model. The model learns motion field adaptions through dedicated MOFA blocks trained on video datasets. It supports multiple control modes including trajectory-based and keypoint-based animation, with training and inference code released publicly on HuggingFace.
Frequently asked
- What is MyNiuuu/MOFA-Video?
- An image-to-video diffusion model that animates static images into videos using learned motion field adaptions.
- Is MOFA-Video open source?
- Yes — MyNiuuu/MOFA-Video is an open-source project tracked on heatdrop.
- What language is MOFA-Video written in?
- MyNiuuu/MOFA-Video is primarily written in Python.
- How popular is MOFA-Video?
- MyNiuuu/MOFA-Video has 765 stars on GitHub.
- Where can I find MOFA-Video?
- MyNiuuu/MOFA-Video is on GitHub at https://github.com/MyNiuuu/MOFA-Video.