Exploring Adversarial Attacks and Defenses in Vision Transformers trained with DINO. (arXiv:2206.06761v2 [cs.CV] UPDATED)

This work conducts the first analysis on the robustness against adversarial
attacks on self-supervised Vision Transformers trained using DINO. First, we
evaluate whether features learned through self-supervision are more robust to
adversarial attacks than those emerging from supervised learning. Then, we
present properties arising for attacks in the latent space. Finally, we
evaluate whether three well-known defense strategies can increase adversarial
robustness in downstream tasks by only fine-tuning the classification head to
provide robustness even in view of limited compute …

arxiv attacks cv transformers vision