Comparison of the Ability of Neural Network Model and Humans to Detect a Cloned Voice

Author:

Milewski Krzysztof1,Zaporowski Szymon12,Czyżewski Andrzej1ORCID

Affiliation:

1. Multimedia Systems Department, Gdansk University of Technology, Narutowicza 11/12, 80-233 Gdańsk, Poland

2. Audio Acoustics Laboratory, Gdansk University of Technology, Narutowicza 11/12, 80-233 Gdańsk, Poland

Abstract

The vulnerability of the speaker identity verification system to attacks using voice cloning was examined. The research project assumed creating a model for verifying the speaker’s identity based on voice biometrics and then testing its resistance to potential attacks using voice cloning. The Deep Speaker Neural Speaker Embedding System was trained, and the Real-Time Voice Cloning system was employed based on the SV2TTS, Tacotron, WaveRNN, and GE2E neural networks. The results of attacks using voice cloning were analyzed and discussed in the context of a subjective assessment of cloned voice fidelity. Subjective test results and attempts to authenticate speakers proved that the tested biometric identity verification system might resist voice cloning attacks even if humans cannot distinguish cloned samples from original ones.

Funder

Polish National Centre for Research and Development (NCBR) from the European Regional Development Fund

Electronics, Telecommunications, and Informatics Faculty, Gdańsk University of Technology

Publisher

MDPI AG

Subject

Electrical and Electronic Engineering,Computer Networks and Communications,Hardware and Architecture,Signal Processing,Control and Systems Engineering

Reference48 articles.

1. (2019, October 11). Available online: https://commonvoice.mozilla.org/en.

2. CorentinJ (2023, January 17). Real Time Voice Cloning. Available online: https://github.com/CorentinJ/Real-Time-Voice-Cloning.

3. Li, C., Ma, X., Jiang, B., Li, X., Zhang, X., Liu, X., Cao, Y., Kannan, A., and Zhu, Z. (2023, January 17). Deep Speaker: An End-to-End Neural Speaker Embedding System. Available online: https://arxiv.org/abs/1705.02304.

4. Liu, X., Sahidullah, M., and Kinnunen, T.A. (2020, January 25–29). Comparative Re-Assessment of Feature Extractors for Deep Speaker Embeddings. Proceedings of the Interspeech, Shanghai, China.

5. Perceptual linear predictive (PLP) analysis of speech;Hermansky;J. Acoust. Soc. Am.,1990

Cited by 3 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

1. Forensic Perspective on Voice Biometrics and AI : A Review;International Journal of Scientific Research in Science and Technology;2024-09-04

2. A computer-aid speech rehabilitation system with mirrored video generating;Technology and Health Care;2024-05-31

3. Using Virtual Reality Technologies to Cope with Grief and Emotional Resilience of Absent Loved Ones;Proceedings of the 2024 8th International Conference on Medical and Health Informatics;2024-05-17

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