PASA: Attack Agnostic Unsupervised Adversarial Detection using
https://arxiv.org:443/abs/2404.10789
WEB6 days ago · Our method, PASA, requires the computation of two test statistics using model prediction and feature attribution and can reliably detect adversarial samples using thresholds learned from benign samples. We validate our lightweight approach by evaluating the performance of PASA on varying strengths of FGSM, PGD, BIM, and CW …
DA: 87 PA: 57 MOZ Rank: 66