kabachuha / sd-webui-text2video

Auto1111 extension implementing text2video diffusion models (like ModelScope or VideoCrafter) using only Auto1111 webui dependencies
Other
1.28k stars 107 forks source link

[Feature Request]: img2vid #44

Closed gsgoldma closed 1 year ago

gsgoldma commented 1 year ago

Is there an existing issue for this?

What would your feature do ?

take an existing image, and make a video that starts with it or incorporates it in some frame

Proposed workflow

  1. Go to ....
  2. Press ....
  3. ...

Additional information

We already have a lot of methods for vid2vid, including using img2img on each image in the video, or pix2pix, or even scripts that do it. We don't have an img2vid yet though. Not sure if this is already in the works.

kabachuha commented 1 year ago

Masking the first frame + denoising rest of the latents sounds like a good solution for this one 👍

Compviztr commented 1 year ago

Can this technique from @alexvw be merged in https://github.com/alexvw/img2video

kabachuha commented 1 year ago

Sounds like a good idea

kabachuha commented 1 year ago

Implemented