Depth-conditional support?

[pj4533]

In the Stable Diffusion v2 repo, it says: "To augment the well-established img2img functionality of Stable Diffusion, we provide a shape-preserving stable diffusion model."

With the current pull request(s) for image2image, is it within scope to add support for shape preserving models?