Pre-trained Text-to-Image Diffusion Models Are Versatile Representation Learners for Control

Devs

Pre-trained Text-to-Image Diffusion Models Are Versatile Representation Learners for Control | Read Paper on Bytez