Improved Speech Authenticity Detection in Chinese–English Bilingual Contexts
Sensors · 2024
DOI: 10.3390/s24216807 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, 2024. Verfasst von Cheng-Yuan Tsai und 4 weiteren Autor:innen. Beteiligte Einrichtung: Forensic Science Division, Ministry of Justice Investigation Bureau, New Taipei City 231, Taiwan. Hauptthemen laut MEDLINE: Language, Multilingualism. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Justice, National Science and Technology Council.
Abstract (Originalsprache)
The rapid evolution of voice technology has heightened the need for robust detection systems to distinguish between authentic and tampered speech. Recent competitions have significantly advanced the development of countermeasures against spoofing attacks. However, while advancements in detection technologies have been notable, existing methods often focus on a single type of tampering and language. Our contribution lies in developing an improved model that integrates an enhanced ResNet architecture with an LSTM to improve the detection of tampered audio, particularly in challenging multilingual scenarios. In the experiments, we built a hybrid dataset from self-recording Chinese speech and public VCTK2 English samples, enhanced the ResNet model generalization capabilities, and evaluated our approach using the bilingual dataset. Experiment results demonstrate that the proposed approach achieves a superior performance with an equal error rate of 11.62%, even in the face of bilingual conditions, and, more importantly, outperforms the leading models from ASVSpoof 2021 and ADD 2022 competitions. We also employed advanced tampering techniques, including CycleGAN voice conversion and auto splicing, to simulate real-world tampering scenarios and verify the effectiveness of the proposed approach.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Forensic Science Division, Ministry of Justice Investigation Bureau, New Taipei City 231, Taiwan.
- Department of Electrical Engineering and AI Research Center, Yuan-Ze University, Taoyuan City 320, Taiwan.
- Department of Electrical Engineering, National Taiwan Normal University, Taipei City 106, Taiwan.
ORCID-Profile: Shih-Hau Fang
Förderung
- Ministry of Justice
- National Science and Technology Council
Zitieren
APA
Tsai, C. Y., Chang, S. C., Hung, C. H., Wang, S. S., & Fang, S. H. (2024). Improved Speech Authenticity Detection in Chinese–English Bilingual Contexts. Sensors. https://doi.org/10.3390/s24216807
Vancouver
Tsai CY, Chang SC, Hung CH, Wang SS, Fang SH. Improved Speech Authenticity Detection in Chinese–English Bilingual Contexts. Sensors. 2024. doi:10.3390/s24216807
BibTeX
@article{Tsai2024,
author = {Tsai, Cheng-Yuan and Chang, Sheng-Chain and Hung, Chao-Hsiang and Wang, Syu-Siang and Fang, Shih-Hau},
title = {Improved Speech Authenticity Detection in Chinese–English Bilingual Contexts},
journal = {Sensors},
year = {2024},
doi = {10.3390/s24216807}
}
EUVIMED