Language-Invariant Multilingual Speaker Verification for the TidyVoice 2026 Challenge

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Li, Ze, Miao, Xiaoxiao, Liu, Juan, Li, Ming
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908873178218496
author Li, Ze
Miao, Xiaoxiao
Liu, Juan
Li, Ming
author_facet Li, Ze
Miao, Xiaoxiao
Liu, Juan
Li, Ming
contents Multilingual speaker verification (SV) remains challenging due to limited cross-lingual data and language-dependent information in speaker embeddings. This paper presents a language-invariant multilingual SV system for the TidyVoice 2026 Challenge. We adopt the multilingual self-supervised w2v-BERT 2.0 model as the backbone, enhanced with Layer Adapters and Multi-scale Feature Aggregation to better exploit multi-layer representations. A language-adversarial training strategy with a Gradient Reversal Layer is applied to promote language-invariant speaker embeddings. Moreover, a multilingual zero-shot text-to-speech system is used to synthesize speech in multiple languages, improving language diversity. Experimental results demonstrate that fine-tuning the large-scale pretrained model yields competitive performance, while language-adversarial training further enhances robustness. In addition, synthetic speech augmentation provides additional gains under limited training data conditions. Source code is available at https://github.com/ZXHY-82/LI-MSV-TidyVoice2026.
format Preprint
id arxiv_https___arxiv_org_abs_2603_08092
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Language-Invariant Multilingual Speaker Verification for the TidyVoice 2026 Challenge
Li, Ze
Miao, Xiaoxiao
Liu, Juan
Li, Ming
Audio and Speech Processing
Multilingual speaker verification (SV) remains challenging due to limited cross-lingual data and language-dependent information in speaker embeddings. This paper presents a language-invariant multilingual SV system for the TidyVoice 2026 Challenge. We adopt the multilingual self-supervised w2v-BERT 2.0 model as the backbone, enhanced with Layer Adapters and Multi-scale Feature Aggregation to better exploit multi-layer representations. A language-adversarial training strategy with a Gradient Reversal Layer is applied to promote language-invariant speaker embeddings. Moreover, a multilingual zero-shot text-to-speech system is used to synthesize speech in multiple languages, improving language diversity. Experimental results demonstrate that fine-tuning the large-scale pretrained model yields competitive performance, while language-adversarial training further enhances robustness. In addition, synthetic speech augmentation provides additional gains under limited training data conditions. Source code is available at https://github.com/ZXHY-82/LI-MSV-TidyVoice2026.
title Language-Invariant Multilingual Speaker Verification for the TidyVoice 2026 Challenge
topic Audio and Speech Processing
url https://arxiv.org/abs/2603.08092