Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors
Abstract
We study the problem of generating adversarial examples in a black-box setting in which only loss-oracle access to a model is available. We introduce a framework that conceptually unifies much of the existing work on black-box attacks, and demonstrate that the current state-of-the-art methods are optimal in a natural sense. Despite this optimality, we show how to improve black-box attacks by bringing a new element into the problem: gradient priors. We give a bandit optimization-based algorithm that allows us to seamlessly integrate any such priors, and we explicitly identify and incorporate two examples. The resulting methods use two to four times fewer queries and fail two to five times less than the current state-of-the-art. The code for reproducing our work is available at https://git.io/fAjOJ.
Cite
Text
Ilyas et al. "Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors." International Conference on Learning Representations, 2019.Markdown
[Ilyas et al. "Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors." International Conference on Learning Representations, 2019.](https://mlanthology.org/iclr/2019/ilyas2019iclr-prior/)BibTeX
@inproceedings{ilyas2019iclr-prior,
title = {{Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors}},
author = {Ilyas, Andrew and Engstrom, Logan and Madry, Aleksander},
booktitle = {International Conference on Learning Representations},
year = {2019},
url = {https://mlanthology.org/iclr/2019/ilyas2019iclr-prior/}
}