正在加载视频...
视频加载失败
Self-Refining Video Sampling: inference-time method using a video generator as its own refiner to correct physics and motion. no retraining needed; scores >70% human preference; is validated on Wan2.2 & Cosmos.
32,230 次观看 • 9 个月前 •via X (Twitter)
3 条评论

Dustin Hollywood9 个月前
💯🔥

nooooob9 个月前
Pipeline level RL-ish impl is a trend

Cube9 个月前
well damn, gonna have to try this out

