arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.05154cs.CV
收藏

基于自适应自动攻击的对抗鲁棒性实用评估

Practical Evaluation of Adversarial Robustness via Adaptive Auto Attack

University of Electronic Science and Technology of China(电子科技大学);Beihang University(北京航空航天大学)

Ye Liu, Yaya Cheng, Lianli Gao, Xianglong Liu, Qilong Zhang, Jingkuan Song

摘要:

针对对抗攻击的防御模型已经显著增长,但缺乏实用的评估方法阻碍了进展。评估可以定义为在给定迭代预算和测试数据集的情况下寻找防御模型的鲁棒性下限。实用的评估方法应该是方便的(即,无参数),有效(即,更少的迭代)和可靠(即,接近鲁棒性的下限)。为此,我们提出了一个无参数的自适应自动攻击(A $^3$)的评估方法,解决了测试时间训练的方式的效率和可靠性。具体来说,通过观察到特定防御模型的对抗性示例在其起点上遵循一些限制,我们设计了自适应方向限制策略来加速评估。此外,为了在迭代预算数下接近鲁棒性的下限,我们提出了一种基于在线算法的丢弃策略,该策略自动识别和丢弃难以攻击的图像。大量的实验证明了我们的A $^3$的有效性。特别地,我们将A $^3$应用于近50种广泛使用的防御模型。通过消耗比现有方法少得多的迭代,即,$1/10$平均(10 $\times$加速),我们在所有情况下实现较低的鲁棒精度。值得注意的是,我们使用此方法在CVPR 2021防御模型白盒对抗攻击比赛中从1681支队伍中赢得了$\textbf{first place}$。代码可从以下网址获得:$\href{https://github.com/liuye6666/adaptive_auto_attack}{https://github.com/liuye6666/adaptive\_auto\_attack}$

英文摘要:

Defense models against adversarial attacks have grown significantly, but the lack of practical evaluation methods has hindered progress. Evaluation can be defined as looking for defense models' lower bound of robustness given a budget number of iterations and a test dataset. A practical evaluation method should be convenient (i.e., parameter-free), efficient (i.e., fewer iterations) and reliable (i.e., approaching the lower bound of robustness). Towards this target, we propose a parameter-free Adaptive Auto Attack (A$^3$) evaluation method which addresses the efficiency and reliability in a test-time-training fashion. Specifically, by observing that adversarial examples to a specific defense model follow some regularities in their starting points, we design an Adaptive Direction Initialization strategy to speed up the evaluation. Furthermore, to approach the lower bound of robustness under the budget number of iterations, we propose an online statistics-based discarding strategy that automatically identifies and abandons hard-to-attack images. Extensive experiments demonstrate the effectiveness of our A$^3$. Particularly, we apply A$^3$ to nearly 50 widely-used defense models. By consuming much fewer iterations than existing methods, i.e., $1/10$ on average (10$\times$ speed up), we achieve lower robust accuracy in all cases. Notably, we won $\textbf{first place}$ out of 1681 teams in CVPR 2021 White-box Adversarial Attacks on Defense Models competitions with this method. Code is available at: $\href{https://github.com/liuye6666/adaptive_auto_attack}{https://github.com/liuye6666/adaptive\_auto\_attack}$

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。

↑