arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.04504cs.LGstat.ML
收藏

诱骗对抗攻击失败

Tricking Adversarial Attacks To Fail

Aalto University(阿尔托大学)

Blerta Lindqvist

摘要:

最近的对抗性防御方法失败了。无针对性的基于梯度的攻击会导致分类器选择任何错误的类。我们新颖的白盒防御技巧使非目标攻击成为针对指定目标类的攻击。从这些目标类中,我们可以派生出真正的类。我们的目标训练防御欺骗了无目标的、基于梯度的对抗攻击核心的最小化:最小化(1)扰动和(2)分类器对抗损失的总和。目标训练最小程度地改变分类器,并使用标记有指定类的额外重复点(距离为0)对其进行训练。这些不同标记的复制样本使最小化的项(1)和(2)两者最小化,从而将攻击收敛引导到指定类别的样本,从中导出正确的分类。重要的是,目标训练消除了了解攻击的需要,以及生成对抗性攻击样本的开销,从而最大限度地减少干扰。我们在CIFAR 10中获得了CW-L2(置信度=0)的86.2%的准确率,甚至超过了非对抗性样本的不安全分类器准确率。目标训练提出了对抗性防御策略的根本变化。

英文摘要:

Recent adversarial defense approaches have failed. Untargeted gradient-based attacks cause classifiers to choose any wrong class. Our novel white-box defense tricks untargeted attacks into becoming attacks targeted at designated target classes. From these target classes, we can derive the real classes. Our Target Training defense tricks the minimization at the core of untargeted, gradient-based adversarial attacks: minimize the sum of (1) perturbation and (2) classifier adversarial loss. Target Training changes the classifier minimally, and trains it with additional duplicated points (at 0 distance) labeled with designated classes. These differently-labeled duplicated samples minimize both terms (1) and (2) of the minimization, steering attack convergence to samples of designated classes, from which correct classification is derived. Importantly, Target Training eliminates the need to know the attack and the overhead of generating adversarial samples of attacks that minimize perturbations. We obtain an 86.2% accuracy for CW-L2 (confidence=0) in CIFAR10, exceeding even unsecured classifier accuracy on non-adversarial samples. Target Training presents a fundamental change in adversarial defense strategy.

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。

↑