omerbt / TokenFlow

Official Pytorch Implementation for "TokenFlow: Consistent Diffusion Features for Consistent Video Editing" presenting "TokenFlow" (ICLR 2024)
https://diffusion-tokenflow.github.io
MIT License
1.52k stars 134 forks source link

Hello, I'd like to ask, does every video require preprocessing before editing? #28

Open a869072989 opened 10 months ago

a869072989 commented 10 months ago

This is an excellent project, but I'm curious, does each video need preprocessing before editing? Thank you very much.

rakesh-reddy95 commented 8 months ago

yes, because you need latents to edit the video. Or you can initiate with random noise which may not reconstruct properly. You can generate the latents ( preprocess) once and edit the video n times with different prompts.