AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems

doi:10.1109/TDSC.2022.3189397

	AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems
	Guangke Chen1,2,3 ; Zhe Zhao1 ; Fu Song1 ; Sen Chen 4; Lingling Fan 5; Yang Liu 6
	2022
发表期刊	IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING (IF:7.0[JCR-2023],6.8[5-Year])
ISSN	2160-9209
卷号	PP 期号:99
发表状态	已发表
DOI	10.1109/TDSC.2022.3189397
摘要	Recent work has illuminated the vulnerability of speaker recognition systems (SRSs) against adversarial attacks, raising significant security concerns in deploying SRSs. However, they considered only a few settings (e.g., some combinations of source and target speakers), leaving many interesting and important settings in real-world attack scenarios alone. In this work, we present AS2T , the first attack in this domain which covers all the settings, thus allows the adversary to craft adversarial voices using arbitrary source and target speakers for any of three main recognition tasks. Since none of the existing loss functions can be applied to all the settings, we explore many candidate loss functions for each setting including the existing and newly designed ones. We thoroughly evaluate their efficacy and find that some existing loss functions are suboptimal. Then, to improve the robustness of AS2T towards practical over-the-air attack, we study the possible distortions occurred in over-the-air transmission, utilize different transformation functions with different parameters to model those distortions, and incorporate them into the generation of adversarial voices. Our simulated over-the-air evaluation validates the effectiveness of our solution in producing robust adversarial voices which remain effective under various hardware devices and various acoustic environments with different reverberation, ambient noises, and noise levels. Finally, we leverage AS2T to perform thus far the largest-scale evaluation to understand transferability among 14 diverse SRSs. The transferability analysis provides many interesting and useful insights which challenge several findings and conclusion drawn in previous works in the image domain. Our study also sheds light on future directions of adversarial attacks in the speaker recognition domain.
URL	查看原文
收录类别	SCI ; SCIE ; EI
来源库	IEEE
文献类型	期刊论文
条目标识符	https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/203460
专题	信息科学与技术学院_博士生信息科学与技术学院_PI研究组_宋富组
作者单位	1.School of Information Science and Technology, ShanghaiTech University, Shanghai, China 2.Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai, China 3.University of Chinese Academy of Sciences, Beijing, China 4.College of Intelligence and Computing, Tianjin University, Tianjin, China 5.College of Cyber Science, Nankai University, Tianjin, China 6.School of Computer Science and Engineering, Nanyang Technological University, Singapore
第一作者单位	信息科学与技术学院
第一作者的第一单位	信息科学与技术学院
推荐引用方式 GB/T 7714	Guangke Chen,Zhe Zhao,Fu Song,et al. AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems[J]. IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING,2022,PP(99).
APA	Guangke Chen,Zhe Zhao,Fu Song,Sen Chen,Lingling Fan,&Yang Liu.(2022).AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems.IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING,PP(99).
MLA	Guangke Chen,et al."AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems".IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING PP.99(2022).