Sparse-RS: a versatile framework for query-efficient sparse black-box adversarial attacks

Code

• github.com/fra31/sparse-rs
  OfficialIn paperpytorch★ 46
• github.com/max-andr/square-attack
  tf★ 169

Abstract

We propose a versatile framework based on random search, Sparse-RS, for score-based sparse targeted and untargeted attacks in the black-box setting. Sparse-RS does not rely on substitute models and achieves state-of-the-art success rate and query efficiency for multiple sparse attack models: l_0-bounded perturbations, adversarial patches, and adversarial frames. The l_0-version of untargeted Sparse-RS outperforms all black-box and even all white-box attacks for different models on MNIST, CIFAR-10, and ImageNet. Moreover, our untargeted Sparse-RS achieves very high success rates even for the challenging settings of 2020 adversarial patches and 2-pixel wide adversarial frames for 224224 images. Finally, we show that Sparse-RS can be applied to generate targeted universal adversarial patches where it significantly outperforms the existing approaches. The code of our framework is available at https://github.com/fra31/sparse-rs.

Tasks

Reproductions