Stable Diffusion is a state of the art text-to-image model that generates images from text. Finetuning the model can make it suitable for inpainting when provided with a starting image, mask, and text prompt.
However, depending on how complex the area you want to mask is, creating the mask can be tedious. This demo incorporates a segmentation model to generate per-class masks for you, which can be combined to produce a final diffusion mask.