Deep neural rejection against adversarial examples

Author:

Sotgiu Angelo,Demontis Ambra,Melis Marco,Biggio Battista,Fumera Giorgio,Feng Xiaoyi,Roli Fabio

Abstract

AbstractDespite the impressive performances reported by deep neural networks in different application domains, they remain largely vulnerable to adversarial examples, i.e., input samples that are carefully perturbed to cause misclassification at test time. In this work, we propose a deep neural rejection mechanism to detect adversarial examples, based on the idea of rejecting samples that exhibit anomalous feature representations at different network layers. With respect to competing approaches, our method does not require generating adversarial examples at training time, and it is less computationally demanding. To properly evaluate our method, we define an adaptive white-box attack that is aware of the defense mechanism and aims to bypass it. Under this worst-case setting, we empirically show that our approach outperforms previously proposed methods that detect adversarial examples by only analyzing the feature representation provided by the output network layer.

Publisher

Springer Science and Business Media LLC

Subject

Computer Science Applications,Signal Processing

Reference40 articles.

1. A. D. Joseph, B. Nelson, B. I. P. Rubinstein, J. Tygar, Adversarial machine learning (Cambridge University Press, 2018).

2. B. Biggio, F. Roli, Wild patterns: ten years after the rise of adversarial machine learning. Pattern. Recog.84:, 317–331 (2018).

3. N Dalvi, P Domingos, Mausam, S Sanghai, D Verma, in Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD). Adversarial classification (Seattle, 2004), pp. 99–108.

4. D. Lowd, C. Meek, in Second Conference on Email and Anti-Spam (CEAS). Good word attacks on statistical spam filters (Mountain ViewUSA, 2005).

5. B. Biggio, B. Nelson, P. Laskov, in 29th Int’l Conf. on Machine Learning, ed. by J. Langford, J. Pineau. Poisoning attacks against support vector machines (Omnipress, 2012), pp. 1807–1814.

Cited by 39 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3