Pre-trained Subject Representation for Controllable Text-to-Image Generation and Editing
Want to make some of these yourself?
