TAPT: Test-Time Adversarial Prompt Tuning for Robust Inference in Vision-Language Models.

Xin Wang, Kai Chen, Jiaming Zhang, Jingjing Chen, Xingjun Ma

Browse the full CVPR paper archive.