By skyforbes My goal is to get depth images of both real and anime characters, as well as gesture sketches. The catch is that I need them in a general form, similar to how NanoBanana can create a simple depth map of a 3D model, without copying the character’s actual pose, especially the foreshortened poses common in anime.
But it isn’t able to do that. And even when it does, the sketch or depth map is about 80 percent off.
I used to try getting prompts from Gemini 3 Pro, but that also failed. I tried explaining it to Gemini the same way I wrote it here, but I’m sure someone has managed to pull this off already.