#168 · Primary category: Video & Animation

ECCV2022-RIFE

aigc computer-vision deep-learning slomo-filter video-interpolation

ECCV2022 - Real-Time Intermediate Flow Estimation for Video Frame Interpolation

Project last updated:09/10/25

GitHub Stars

5.6K

Forks

564

Contributors

24

License

MIT

Why we included this project

If you edit video and need frames that never existed in the footage, RIFE is a solid place to start. It estimates intermediate optical flow between two frames and can place the interpolated frame at any timestep, so the same model handles doubling 24fps footage to 48 or 60fps and building slow motion up to 16x from a pair of images. The original release runs at 30+ FPS for 2x 720p interpolation on a 2080Ti, fast enough for interactive use on mid-range hardware, and later weights add models tuned for anime scenes alongside photo-real ones. The repo ships command-line scripts that feed frames through ffmpeg into a finished mp4 or gif. Since the authors also link third-party GUIs like Flowframes and SVFI plus a Colab demo, you can try it without writing code and move to your own pipeline when you need it.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category