semantic image editing

[ljleb]

Last year a paper and this repo have been published: https://github.com/ml-research/semantic-image-editing

If I understand correctly this allows to edit images with a series of editing prompts. Is there a reason I cannot find any related discussion in the webui repo? Is there a reason it has not yet been implemented either as a builtin feature or as an extension?

Edit: I don't say this to pressure implementation, but to know whether there are pitfalls to this technique and whether it was considered before.

[ljleb]

Hmm but now that pix2pix is possible in the webui, I'm not sure this is useful anymore.