2024 09_eccv

Adversarial Robustification via Text-to-Image Diffusion Models selected for Oral Presentation (top 2.3%) at ECCV 2024.