[Feature Request]: Please add inpainting frames to vid2vid

kabachuha / sd-webui-text2video

Auto1111 extension implementing text2video diffusion models (like ModelScope or VideoCrafter) using only Auto1111 webui dependencies

Other

1.28k stars 106 forks source link

[Feature Request]: Please add inpainting frames to vid2vid #147

Open alexfredo opened 1 year ago

alexfredo commented 1 year ago

Is there an existing issue for this?

[X] I have searched the existing issues and checked the recent builds/commits

What would your feature do ?

Hi it's could be awesome if vid2vid have "img2vid inpainting frames" for help to make transition between each part of a video

Proposed workflow

Add img2vid in the vid2vid tab

Additional information

No response

alexfredo commented 1 year ago

Also it's could be automated with a system that automatically keep the last generated frame and use it for generate the next video, for example Im limited to 19 frames at 512x512 every 19 frames it stop and restart the processus with the last frame, for make very long video with low vram