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 2103.11526cs.LG
收藏

ExAD:一种基于包围圈的对抗检测方法

ExAD: An Ensemble Approach for Explanation-based Adversarial Detection

Texas A&M University(德克萨斯农工大学);Hefei University of Technology(合肥工业大学)

Raj Vardhan, Ninghao Liu, Phakpoom Chinprutthiwong, Weijie Fu, Zhenyu Hu, Xia Ben Hu, Guofei Gu

摘要:

最近的研究表明,深度神经网络(DNN)容易受到对抗性示例的影响,这些示例会在模型中引起所需的错误分类。这些风险阻碍了机器学习在安全敏感领域的应用。针对对抗攻击,已经提出了几种防御方法,以在测试时检测对抗示例或使机器学习模型更鲁棒。然而,虽然现有的方法在黑盒威胁模型下是非常有效的,其中攻击者不知道防御,但它们在白盒威胁模型下相对无效,其中攻击者完全知道防御。在本文中,我们提出了ExAD,这是一个使用解释技术集成来检测对抗性示例的框架。ExAD中的每种解释技术都会产生一个解释图,用于识别模型分类的输入变量的相关性。对于数据集中的每个类,系统包括与每种解释技术相对应的检测器网络,该检测器网络经过训练以区分正常和异常解释图。在测试时,如果输入的解释图被分类类的任何检测器模型检测为异常,则我们认为输入是对抗性示例。我们使用六个最先进的对抗攻击对三个图像数据集评估我们的方法。我们的广泛的评估表明,我们的机制可以有效地检测这些攻击的黑盒威胁模型有限的误报。此外,我们发现,我们的方法在限制白盒攻击的成功率方面取得了可喜的成果。

英文摘要:

Recent research has shown Deep Neural Networks (DNNs) to be vulnerable to adversarial examples that induce desired misclassifications in the models. Such risks impede the application of machine learning in security-sensitive domains. Several defense methods have been proposed against adversarial attacks to detect adversarial examples at test time or to make machine learning models more robust. However, while existing methods are quite effective under blackbox threat model, where the attacker is not aware of the defense, they are relatively ineffective under whitebox threat model, where the attacker has full knowledge of the defense. In this paper, we propose ExAD, a framework to detect adversarial examples using an ensemble of explanation techniques. Each explanation technique in ExAD produces an explanation map identifying the relevance of input variables for the model's classification. For every class in a dataset, the system includes a detector network, corresponding to each explanation technique, which is trained to distinguish between normal and abnormal explanation maps. At test time, if the explanation map of an input is detected as abnormal by any detector model of the classified class, then we consider the input to be an adversarial example. We evaluate our approach using six state-of-the-art adversarial attacks on three image datasets. Our extensive evaluation shows that our mechanism can effectively detect these attacks under blackbox threat model with limited false-positives. Furthermore, we find that our approach achieves promising results in limiting the success rate of whitebox 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。

↑