arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.09172cs.LGcs.CR
收藏

可转移对抗样本的方向聚集攻击

Direction-Aggregated Attack for Transferable Adversarial Examples

Eindhoven University of Technology(埃因霍温理工大学);National University of Singapore(新加坡国立大学)

Tianjin Huang, Vlado Menkovski, Yulong Pei, YuHao Wang, Mykola Pechenizkiy

摘要:

深度神经网络很容易受到对抗性示例的攻击,这些示例是通过对输入进行难以察觉的更改而制作的。然而,这些对抗性示例在白盒设置中最成功,其中模型及其参数可用。找到可转移到其他模型或在黑盒环境中开发的对抗性示例要困难得多。在本文中,我们提出了方向聚合对抗攻击,提供可转移的对抗示例。我们的方法在攻击过程中利用聚合方向,以避免生成的对抗性示例过度拟合白盒模型。在ImageNet上进行的大量实验表明,我们提出的方法显著提高了对抗性示例的可移植性,并且优于最先进的攻击,特别是对抗性鲁棒模型。该方法对三种对抗训练模型的最佳平均攻击成功率为94.6%,对五种防御方法的最佳平均攻击成功率为94.8%.它还揭示了当前的防御方法不能防止可转移的对抗攻击。

英文摘要:

Deep neural networks are vulnerable to adversarial examples that are crafted by imposing imperceptible changes to the inputs. However, these adversarial examples are most successful in white-box settings where the model and its parameters are available. Finding adversarial examples that are transferable to other models or developed in a black-box setting is significantly more difficult. In this paper, we propose the Direction-Aggregated adversarial attacks that deliver transferable adversarial examples. Our method utilizes aggregated direction during the attack process for avoiding the generated adversarial examples overfitting to the white-box model. Extensive experiments on ImageNet show that our proposed method improves the transferability of adversarial examples significantly and outperforms state-of-the-art attacks, especially against adversarial robust models. The best averaged attack success rates of our proposed method reaches 94.6\% against three adversarial trained models and 94.8\% against five defense methods. It also reveals that current defense approaches do not prevent transferable adversarial attacks.

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。

↑