One Prompt Word is Enough to Boost Adversarial Robustness for Pre-trained Vision-Language Models
暂无评论，来聊聊你的看法吧