作者位次 #6/8ICML 2026Poster· Deep Learning / Robustness
Two Modalities Are Better Than One: Efficient Adversarial Purification via Multimodal Diffusion Models
Mingyuan Bai · Wei Huang · Tenghui Li · Andong Wang · Chao Li · Cesar F Caiafa · Junbin Gao · Qibin Zhao