arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.05040cs.LGcs.CRcs.CVstat.ML
收藏

稀疏和不可感知的对抗攻击

Sparse and Imperceivable Adversarial Attacks

University of Tübingen(图宾根大学)

Francesco Croce, Matthias Hein

摘要:

神经网络已被证明易受各种对抗攻击的攻击。从安全的角度来看,高度稀疏的对抗攻击尤其危险。另一方面,稀疏攻击的像素级扰动通常很大,因此可以潜在地检测到。我们提出了一种新的黑盒技术来制作对抗性的例子,旨在最大限度地减少$l_0$距离原始图像。大量的实验表明,我们的攻击是更好的或竞争的最先进的。此外,我们可以集成额外的边界上的分量扰动。允许像素仅在高变化区域中变化,并避免沿轴对齐边缘的变化,这使得我们的对抗性示例几乎不可感知。此外,我们适应投影梯度下降攻击的$l_0$ -范数集成组件的约束。这使我们能够进行对抗训练,以增强分类器对稀疏和不可感知的对抗操作的鲁棒性。

英文摘要:

Neural networks have been proven to be vulnerable to a variety of adversarial attacks. From a safety perspective, highly sparse adversarial attacks are particularly dangerous. On the other hand the pixelwise perturbations of sparse attacks are typically large and thus can be potentially detected. We propose a new black-box technique to craft adversarial examples aiming at minimizing $l_0$-distance to the original image. Extensive experiments show that our attack is better or competitive to the state of the art. Moreover, we can integrate additional bounds on the componentwise perturbation. Allowing pixels to change only in region of high variation and avoiding changes along axis-aligned edges makes our adversarial examples almost non-perceivable. Moreover, we adapt the Projected Gradient Descent attack to the $l_0$-norm integrating componentwise constraints. This allows us to do adversarial training to enhance the robustness of classifiers against sparse and imperceivable adversarial manipulations.

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。

↑