Fast and Reliable Evaluation of Adversarial Robustness with Minimum-Margin Attack. (arXiv:2206.07314v1 [cs.LG])

Web: http://arxiv.org/abs/2206.07314

The AutoAttack (AA) has been the most reliable method to evaluate adversarial
robustness when considerable computational resources are available. However,
the high computational cost (e.g., 100 times more than that of the project
gradient descent attack) makes AA infeasible for practitioners with limited
computational resources, and also hinders applications of AA in the adversarial
training (AT). In this paper, we propose a novel method, minimum-margin (MM)
attack, to fast and reliably evaluate adversarial robustness. Compared with AA,
our method achieves …

arxiv evaluation lg robustness