how to make the target have the context texture in semantic diffusion?

[molo32]

how to make the target have the context texture in semantic diffusion?

in semantic diffusion the "target" changes the texture or skin of the "context" how to avoid that?

I want semantic diffusion to have the target face but the context skin.

[ShenYujun]

Diffusion means to diffuse the target to surroundings. Hence, the results will be with the target skin.

[molo32]

Got it, but do you know any way to edit the code to have the target face but the context skin. style mix does not give good results

[ShenYujun]

Maybe you can try training a style transfer model ;)