What is the specific difference between controlnet_inpainting and sd_inpainting?

[cb614611757]

According to the effect of sd_inpaint, mask_image is used to control the area that needs to be regenerated, so the difficulty is actually the generation of mask image. From the input of controlnet_inpainting and the final effect, it requires an additional seg_map, and the seg_map image does not show its role in the final effect.

The top image is the effect of stable diffusion inpainting and the bottom image is the effect of controlnet_inpainting.

[haofanwang]

For this simple example, yes, they are almost the same, as the segmentation map doesn't provide strong guidance. But if you try to mask a person, and wanna generate a new person with specific pose, a control hint is needed. @cb614611757