Combining Depth, Canny, OpenPose and/or Scribble models [IMPLEMENTED!]

#12
by Sub-Zero - opened

For characters, I see that there are some good advantages for each model, but I think that combining several models/techniques at the same time would probably have much better results.

For example, combining Depth, Canny, OpenPose and/or Scribble (instead of an outline, it would ideally get an image containing simple lines indicating visible character bones [edit: for finger bones]) and taking into consideration several factors at the same time, ControlNet would considerably improve image outputs.

So, the question is, is it possible to combine these ControlNet models ? Thanks :)

Sub-Zero changed discussion title from Combining Depth, Canny, OpenPose and/or Scribble for better humans to Combining Depth, Canny, OpenPose and/or Scribble models

agreed -- would be great to be able to flexible compose a scene using different methods of control!

Sub-Zero changed discussion title from Combining Depth, Canny, OpenPose and/or Scribble models to Combining Depth, Canny, OpenPose and/or Scribble models [IMPLEMENTED!]

This has been implemented!

  1. Update your extension to the latest version.
  2. Go to settings > ControlNet > Multi ControlNet: Max models amount (requires restart) and choose the number of models you want to use at the same time (1 to 10!).
  3. Save settings.

You're a beast, thanks for the quick update! Can't wait to play with multi-control net to do more video2video stuff like this: https://www.youtube.com/watch?v=n1PlJwwKb1k

Sign up or log in to comment