构建跨语言语音识别系统，提升多语言说话人验证效果。
Language-Invariant Multilingual Speaker Verification for the TidyVoice 2026 Challenge
- 用多语言自监督模型+层适配器增强表示能力
- 对抗训练使说话人特征摆脱语言依赖，提升鲁棒性
- 合成语音扩充数据，适合小样本场景
多语言说话人验证（SV）因跨语言数据有限及说话人嵌入中的语言依赖信息而面临挑战。本文为TidyVoice 2026挑战赛提出一种语言不变的多语言说话人验证系统。采用多语言自监督w2v-BERT 2.0作为主干网络，结合层适配器与多尺度特征聚合以更好利用多层表示。引入带有梯度反转层的语言对抗训练策略，促进语言不变的说话人嵌入。此外，使用多语言零样本文本到语音系统合成多种语言语音，增强语言多样性。实验表明，微调大规模预训练模型可取得具有竞争力的性能，而语言对抗训练进一步提升鲁棒性；在训练数据有限时，合成语音增强带来额外增益。源代码已公开于 https://github.com/ZXHY-82/LI-MSV-TidyVoice2026。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。