EvoBA: An Evolution Strategy as a Strong Baseline forBlack-Box Adversarial Attacks

1 code implementation • 12 Jul 2021 • Andrei Ilie, Marius Popescu, Alin Stefanescu

We propose $\textbf{EvoBA}$ as a query-efficient $L_0$ black-box adversarial attack which, together with the aforementioned methods, can serve as a generic tool to assess the empirical robustness of image classifiers.

Adversarial Attack