arrow
返回

Low frequency sparse adversarial attack

delete2023-09-01
delete2
PRE
AI
J
Jiyuan Liu
B
Bingyi Lu
M
Mingkang Xiong
张涛 封面图
张涛 (Tao Zhang)
H
Huilin Xiong *
DOI:10.1016/j.cose.2023.103379delete
delete原文链接
delete分享
delete收藏
摘要

摘要

En 中文
Deep neural network (DNN) is found to be vulnerable to adversarial attacks. Gradient-based adversarial attacks usually craft perturbation on the whole pixels of input image, making them easier to be perceived or detected by the detection-based defenses, in the case of using large perturbation budget. In this paper, we present a new sparse approach to perturb image, which we call Robust Gradient-based Low Frequency search (RGLF). We show that RGLF provides a convenient way to modify any adversarial attack so that its imperceptibility can be improved significantly, not only to human eyes, but also to some popular detection-based defenses. Technically, RGLF first uses a robust gradient mask (RGM) to localize the semantic regions of input image, and then, applies a technique, called adaptive frequency search (AFS), to generate the low-frequency perturbations we expect, and finally, embeds the perturbations in the semantic regions. Extensive experiments are conducted to show that, i) the proposed RGLF technique can remarkably reduce the detection rates of the gradient-based adversarial attacks by 25%, 36%, and 30%, respectively for three SOTA detection-based defenses, namely, SBD, LiBRe, and SimCat; ii) The imperceptibility of the adversarial attacks modified by RGLF is improved significantly, even in the case of using large perturbation budget.
Keyword:
Adversarial attack
Imperceptible attack
Sparse attack
Low-frequency attack
Image classification

期刊

C
Computers and Security
IF:
5.4
论文数:
4.6K
被引数:
1.8W

机构

上海交通大学 封面图
上海交通大学
学者数:
1.3W
论文数: 3.9K
被引数: 8.7W
引用论文

引用论文

Procedure for loading diamond cells with high-pressure gas
err1980-07-01
err0
PREAI
errR. L. Mills; D. H. Liebenberg; J. C. Bronson; L. C. Schmidt
err分享
err收藏
Boosting adversarial attacks with transformed gradient用变换的梯度增强对抗攻击
err2022-07-01
err12
PREAI
errHe, Zhengyun; Duan, Yexin; Zhang, Wu; Zou, Junhua; He, Zhengfang; Wang, Yunyun; Pan, Zhisong
err分享
err收藏
Efficient Synthesis of the Immunomodulating Compound KRP-203
err2007-07-01
err0
PREAI
errMaiko Hamada; Masatoshi Kiuchi; Kunitomo Adachi
err分享
err收藏
Self‐reported child maltreatment and cardiometabolic risk in 30‐year‐old adults
err2022-07-12
err0
errOAAI
errStephen Kisely; Dan Siskind; James G. Scott; Jake M. Najman
err分享
err收藏
学者 查看更多内容