Localizing Object-level Shape Variations with Text-to-Image Diffusion Models

Open in new window