arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.12133cs.LGstat.ML
收藏

可证明鲁棒对抗示例

Provably Robust Adversarial Examples

ETH Zurich(苏黎世联邦理工学院);University of Illinois, Urbana Champaign(伊利诺伊大学厄巴纳-香槟分校);VMware Research(VMware研究院)

Dimitar I. Dimitrov, Gagandeep Singh, Timon Gehr, Martin Vechev

摘要:

我们引入了深度神经网络的可证明鲁棒对抗性示例的概念-从标准对抗性示例构建的连接输入区域,保证对一组真实世界的扰动(例如像素强度和几何变换的变化)具有鲁棒性。我们提出了一种新的方法,称为PARADE生成这些区域的可扩展的方式,其工作原理是通过迭代细化最初通过采样获得的区域,直到一个细化的区域被证明是对抗与现有的国家的最先进的验证。在每一步中,一个新的优化过程被应用到最大限度地提高该区域的体积的约束下,相对于该区域的网络行为的凸松弛意味着一个选定的约束上的认证目标。我们的实验评估表明了PARADE的有效性:它成功地找到了大型可证明的鲁棒区域,包括包含像素强度的$\approx 10^{573}$对抗性示例和几何扰动的$\approx 10^{599}$的区域。可证明性使我们的强大的例子是显着更有效地对国家的最先进的防御基于随机平滑比用于构建区域的个人攻击。

英文摘要:

We introduce the concept of provably robust adversarial examples for deep neural networks - connected input regions constructed from standard adversarial examples which are guaranteed to be robust to a set of real-world perturbations (such as changes in pixel intensity and geometric transformations). We present a novel method called PARADE for generating these regions in a scalable manner which works by iteratively refining the region initially obtained via sampling until a refined region is certified to be adversarial with existing state-of-the-art verifiers. At each step, a novel optimization procedure is applied to maximize the region's volume under the constraint that the convex relaxation of the network behavior with respect to the region implies a chosen bound on the certification objective. Our experimental evaluation shows the effectiveness of PARADE: it successfully finds large provably robust regions including ones containing $\approx 10^{573}$ adversarial examples for pixel intensity and $\approx 10^{599}$ for geometric perturbations. The provability enables our robust examples to be significantly more effective against state-of-the-art defenses based on randomized smoothing than the individual attacks used to construct the regions.

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。

↑