Self-distillation-based domain exploration for source speaker verification under spoofed speech from unknown voice conversion

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F26%3A0201395" target="_blank" >RIV/00216305:26230/26:0201395 - isvavai.cz</a>

Result on the web

<a href="https://www.sciencedirect.com/science/article/pii/S0167639324001249?pes=vor&utm_source=scopus&getft_integrator=scopus" target="_blank" >https://www.sciencedirect.com/science/article/pii/S0167639324001249?pes=vor&utm_source=scopus&getft_integrator=scopus</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.specom.2024.103153" target="_blank" >10.1016/j.specom.2024.103153</a>

Result language

angličtina

Original language name

Self-distillation-based domain exploration for source speaker verification under spoofed speech from unknown voice conversion

Original language description

Advancements in voice conversion (VC) technology have made it easier to generate spoofed speech that closely resembles the identity of a target speaker. Meanwhile, verification systems within the realm of speech processing are widely used to identify speakers. However, the misuse of VC algorithms poses significant privacy and security risks by potentially deceiving these systems. To address this issue, source speaker verification (SSV) has been proposed to verify the source speaker's identity of the spoofed speech generated by VCs. Nevertheless, SSV often suffers severe performance degradation when confronted with unknown VC algorithms, which is usually neglected by researchers. To deal with this cross-voice-conversion scenario and enhance the model's performance when facing unknown VC methods, we redefine it as a novel domain adaptation task by treating each VC method as a distinct domain. In this context, we propose an unsupervised domain adaptation (UDA) algorithm termed self-distillation-based domain exploration (SDDE). This algorithm adopts a siamese framework with two branches: one trained on the source (known) domain and the other trained on the target domains (unknown VC methods). The branch trained on the source domain leverages supervised learning to capture the source speaker's intrinsic features. Meanwhile, the branch trained on the target domain employs self-distillation to explore target domain information from multi-scale segments. Additionally, we have constructed a large-scale data set comprising over 7945 h of spoofed speech to evaluate the proposed SDDE. Experimental results on this data set demonstrate that SDDE outperforms traditional UDAs and substantially enhances the performance of the SSV model under unknown VC scenarios. The code for data generation and the trial lists are available at https://github.com/zrtlemontree/cross-domain-source-speaker-verification.

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

S - Specificky vyzkum na vysokych skolach

Publication year

2025

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Speech communication

ISSN

0167-6393

e-ISSN

1872-7182

Volume of the periodical

167

Issue of the periodical within the volume

103153

Country of publishing house

NL - THE KINGDOM OF THE NETHERLANDS

Number of pages

12

Pages from-to

1-12

UT code for WoS article

001391212500001

EID of the result in the Scopus database

2-s2.0-85212173287