arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.02014cs.LGcs.CRcs.DS
收藏

在线对抗攻击

Online Adversarial Attacks

Mila(米拉人工智能研究所);Université de Montréal(蒙特利尔大学);McGill University(麦吉尔大学);Meta AI Research(Meta人工智能研究院)

Andjela Mladenovic, Avishek Joey Bose, Hugo Berard, William L. Hamilton, Simon Lacoste-Julien, Pascal Vincent, Gauthier Gidel

摘要:

对抗攻击暴露了深度学习模型的重要漏洞,但很少有人关注数据作为流到达的设置。在本文中,我们形式化的在线对抗攻击问题,强调在现实世界的用例中发现的两个关键要素:攻击者必须在目标模型的部分知识下操作,攻击者所做的决定是不可撤销的,因为他们对瞬态数据流进行操作。我们首先严格分析了一个确定性的变体的在线威胁模型绘制平行的理论计算机科学中研究充分的$k$秘书问题,并提出虚拟+,一个简单而实用的在线算法。我们的主要理论结果表明,虚拟+产量可证明是最好的竞争比在所有单阈值算法的$k<5$-扩展以前的分析$k$秘书问题。我们还介绍了\textit{stochastic $k$-secretary} -有效地减少在线黑盒传输攻击的$k$秘书的问题下的噪音-并证明了虚拟+适应这种设置的性能的理论界限。最后,我们通过对MNIST,CIFAR-10和Imagenet分类器进行实验来补充我们的理论结果,揭示在线算法在实现接近最佳性能方面的必要性,以及攻击策略和在线攻击选择之间的丰富相互作用,使FGSM等简单策略能够胜过更强大的对手。

英文摘要:

Adversarial attacks expose important vulnerabilities of deep learning models, yet little attention has been paid to settings where data arrives as a stream. In this paper, we formalize the online adversarial attack problem, emphasizing two key elements found in real-world use-cases: attackers must operate under partial knowledge of the target model, and the decisions made by the attacker are irrevocable since they operate on a transient data stream. We first rigorously analyze a deterministic variant of the online threat model by drawing parallels to the well-studied $k$-secretary problem in theoretical computer science and propose Virtual+, a simple yet practical online algorithm. Our main theoretical result shows Virtual+ yields provably the best competitive ratio over all single-threshold algorithms for $k<5$ -- extending the previous analysis of the $k$-secretary problem. We also introduce the \textit{stochastic $k$-secretary} -- effectively reducing online blackbox transfer attacks to a $k$-secretary problem under noise -- and prove theoretical bounds on the performance of Virtual+ adapted to this setting. Finally, we complement our theoretical results by conducting experiments on MNIST, CIFAR-10, and Imagenet classifiers, revealing the necessity of online algorithms in achieving near-optimal performance and also the rich interplay between attack strategies and online attack selection, enabling simple strategies like FGSM to outperform stronger adversaries.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑