PangzeCheung/OmniTransfer

[CVPR 2026] OmniTransfer: All-in-one Framework for Spatio-temporal Video Transfer

42
/ 100
Emerging

This project helps video creators transform existing video footage by applying various visual characteristics from other reference videos. You can take an input video or image and a reference video, and it generates a new video where the input adopts the visual style, motion, camera movement, or even character identity from the reference. It's for visual effects artists, content creators, or video editors looking to quickly customize and enhance videos with specific artistic or dynamic elements.

224 stars.

Use this if you need to transfer complex visual elements like special effects, character motion, camera work, or unique styles from one video to another without detailed manual editing or pose extraction.

Not ideal if you need frame-by-frame precision for highly sensitive commercial work or if your primary goal is simple video editing like cuts and transitions rather than advanced visual transformation.

video-editing visual-effects content-creation motion-graphics digital-art
No Package No Dependents
Maintenance 10 / 25
Adoption 10 / 25
Maturity 13 / 25
Community 9 / 25

How are scores calculated?

Stars

224

Forks

9

Language

License

Apache-2.0

Last pushed

Feb 21, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/PangzeCheung/OmniTransfer"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.