2024 ECCV ECCV 2024

Free-ATM: Harnessing Free Attention Masks for Representation Learning on Diffusion-Generated Images