AED-PADA:Improving Generalizability of Adversarial Example Detection via Principal Adversarial Domain Adaptation

arXiv:2404.12635v1 Announce Type: cross
Abstract: Adversarial example detection, which can be conveniently applied in many scenarios, is important in the area of adversarial defense. Unfortunately, existing detection methods suffer from poor generalization performance, because their training process usually relies on the examples generated from a single known adversarial attack and there exists a large discrepancy between the training and unseen testing adversarial examples. To address this issue, we propose a novel method, named Adversarial Example Detection via Principal Adversarial Domain …

adaptation adversarial adversarial attack area arxiv attack can cs.cr cs.cv cs.lg defense detection domain examples generated important performance poor process single training