Is a "specific canva space" controlnet planned ?

Nobody has claimed this yet.

I want to push Stable Diffusion utility further, and render pictures out of really specific canvas/ mask size.

Something like this for instance [
](url)


This is just a quick test on photoshop, I know that controlnet segmentation allows some rendering freedom, but you can't really draw anything with it, here is a test:

Mask & Segmentation end up with a visually interesting result, but that is unusable within the mask, what I would like is Stable diffusion to paint whatever I've prompted only inside the white part of the mask, for instance here, having a whole egyptian landscape directly inside the shoe
Similarly to this quick render that I've masked inside the sneaker in photoshop, although there is no real composition regarding the shape of the mask.

Is there any ways to achieve it ?

The issue does not name a file, test, or entry point. Start by determining whether ControlNet supports restricting prompted generation to a specific canvas or mask region, then define the required behavior and validation for an implementation.

Written by the indexing model from the issue text.