English

应用科学学报 ›› 2024, Vol. 42 ›› Issue (5): 782-794.doi: 10.3969/j.issn.0255-8297.2024.05.006

• 信号与信息处理 • 上一篇    

基于深度声纹特征转换网络的说话人识别攻击方法

陶子钰1, 苏兆品1,3,4, 廉晨思2,4, 王年松2,4, 张国富1,3,4   

  1. 1. 合肥工业大学 计算机与信息学院, 安徽 合肥 230009;
    2. 安徽省公安厅 物证鉴定管理处, 安徽 合肥 230000;
    3. 合肥工业大学 智能互联系统安徽省实验室, 安徽 合肥 230009;
    4. 音视频智能防识联合实验室, 安徽 合肥 230009
  • 收稿日期:2023-11-08 发布日期:2024-09-29
  • 通信作者: 陶子钰，研究方向为说话人识别攻击。E-mail:2021111113@hfut.edu.cn E-mail:2021111113@hfut.edu.cn
  • 基金资助:
    安徽省重点研究与开发计划（No.202104d07020001）；安徽省自然科学基金（No.2208085MF166）资助

Attack Towards Speaker Identification Using Deep Conversion Networks for Voiceprint Features

TAO Ziyu1, SU Zhaopin1,3,4, LIAN Chensi2,4, WANG Niansong2,4, ZHANG Guofu1,3,4   

  1. 1. School of Computer and Information Technology, Hefei University of Technology, Hefei 230009, Anhui, China;
    2. Department of Physical Evidence Identification, Anhui Public Security Department, Hefei 230000, Anhui, China;
    3. Intelligent Interconnected Systems Laboratory of Anhui Province, Hefei University of Technology, Hefei 230009, Anhui, China;
    4. Joint Laboratory of Intelligent Prevention and Recognition of Audio and Video, Hefei 230009, Anhui, China
  • Received:2023-11-08 Published:2024-09-29

PDF

88

摘要： 目前主流说话人识别（speaker identification,SID）系统的攻击方法主要基于快速梯度下降或映射式梯度下降算法，这些方法存在攻击效果不稳定、生成的攻击语音听觉质量不高等问题。为此提出一种基于深度声纹特征转换网络的自动说话人识别攻击方法，生成具有目标说话人音色的攻击语音。首先分析了SID系统的攻击流程，确定了攻击语音生成的过程；然后基于二维卷积神经网络设计攻击音频生成器，以有效融合源说话人的语音内容和目标说话人的声纹特征，并基于对抗学习设计了攻击音频的判别器，以提高语音攻击音频的质量。最后分别在基于广义端到端损失和基于AMSoftmax损失的两个自动说话人识别系统上进行对比实验。实验结果表明，所提方法不但提高了攻击效果的稳定性，提升了攻击音频的人耳感受质量，而且适用于短时长数据，满足了实际攻击场景的需求。

关键词: 说话人识别, 攻击语音, 声纹特征转换, 卷积神经网络

Abstract: In the field of speaker identification (SID) systems, attacks often rely on fast gradient descent and mapping gradient descent algorithms, which suffer from unstable attack performance and poor auditory quality of generated attack samples. This paper proposes an advanced attack method against SID systems using deep neural networks to generate attack speeches with the target speaker’s voiceprint. Specifically, the attack process on SID system is first analyzed to determine the approach to generating attack speeches. Then, a two-dimensional convolutional neural network is designed as a generator to effectively integrate the speech content of the source speaker and the voiceprint features of the target speaker. A discriminator is designed based on adversarial learning to improve the quality of the attack speeches. Finally, comparative experiments are conducted on two automatic SID systems based on generalized end-to-end loss and AMSoftmax loss, respectively. Experimental results demonstrate that the proposed method not only improves the stability of attack performance, but also enhances the auditory quality of attack speeches. Moreover, the proposed method is applicable to short samples, making it suitable for practical attack scenarios.

Key words: speaker identification, attack speeches, voiceprint feature conversion, convolutional neural network

中图分类号: