Learning to Fool the Speaker Recognition

doi:10.1145/3468673

	Learning to Fool the Speaker Recognition
	Li, Jiguo 1,2; Zhang, Xinfeng 3; Xu, Jizheng 4; Ma, Siwei 5; Gao, Wen 5
	2021-10-01
发表期刊	ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
ISSN	1551-6857
卷号	17 期号:3 页码:21
摘要	Due to the widespread deployment of fingerprint/face/speaker recognition systems, the risk in these systems, especially the adversarial attack, has drawn increasing attention in recent years. Previous researches mainly studied the adversarial attack to the vision-based systems, such as fingerprint and face recognition. While the attack for speech-based systems has not been well studied yet, although it has been widely used in our daily life. In this article, we attempt to fool the state-of-the-art speaker recognition model and present speaker recognition attacker, a lightweight multi-layer convolutional neural network to fool the well-trained state-of-the-art speaker recognition model by adding imperceptible perturbations onto the raw speech waveform. We find that the speaker recognition system is vulnerable to the adversarial attack, and achieve a high success rate on both the non-targeted attack and targeted attack. Besides, we present an effective method by leveraging a pretrained phoneme recognition model to optimize the speaker recognition attacker to obtain a tradeoff between the attack success rate and the perceptual quality. Experimental results on the TIMIT and LibriSpeech datasets demonstrate the effectiveness and efficiency of our proposed model. And the experiments for frequency analysis indicate that high-frequency attack is more effective than low-frequency attack, which is different from the conclusion drawn in previous image-based works. Additionally, the ablation study gives more insights into our model.
关键词	Audio forensics adversarial attack deep neural network
DOI	10.1145/3468673
收录类别	SCI
语种	英语
资助项目	National Science Foundation of China[62025101620] ; National Science Foundation of China[61961130392] ; PKU-Baidu Fund[2019BD003] ; High-performance Computing Platform of Peking University
WOS研究方向	Computer Science
WOS类目	Computer Science, Information Systems ; Computer Science, Software Engineering ; Computer Science, Theory & Methods
WOS记录号	WOS:000738280600011
出版者	ASSOC COMPUTING MACHINERY
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/18369
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Ma, Siwei
作者单位	1.Chinese Acad Sci, Inst Comp Technol, 6 Kexueyuan South Rd, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, 6 Kexueyuan South Rd, Beijing 100190, Peoples R China 3.Univ Chinese Acad Sci, 19A Yuquan Rd, Beijing 100049, Peoples R China 4.Bytedance Inc, 48A Zhichun Rd, Beijing 100191, Peoples R China 5.Peking Univ, 5 Yiheyuan Rd, Beijing 100871, Peoples R China
推荐引用方式 GB/T 7714	Li, Jiguo,Zhang, Xinfeng,Xu, Jizheng,et al. Learning to Fool the Speaker Recognition[J]. ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS,2021,17(3):21.
APA	Li, Jiguo,Zhang, Xinfeng,Xu, Jizheng,Ma, Siwei,&Gao, Wen.(2021).Learning to Fool the Speaker Recognition.ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS,17(3),21.
MLA	Li, Jiguo,et al."Learning to Fool the Speaker Recognition".ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS 17.3(2021):21.