arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2412.04425eess.AScs.CLcs.LGcs.SD
收藏

CA-SSLR:面向通用语音处理的条件感知自监督学习表示

CA-SSLR: Condition-Aware Self-Supervised Learning Representation for Generalized Speech Processing

Johns Hopkins University(约翰斯·霍普金斯大学);Center for Language and Speech Processing(语言与语音处理中心)

Yen-Ju Lu, Jing Liu, Thomas Thebaud, Laureano Moro-Velazquez, Ariya Rastrow, Najim Dehak, Jesus Villalba

AI总结:

本文提出CA-SSLR通用条件化模型,通过融合早期语言与说话人嵌入并以线性调制动态调整表示,在减少参数和过拟合的同时显著提升多语言语音识别、语种识别和说话人验证性能。

摘要:

我们提出了条件感知自监督学习表示(Condition-Aware Self-Supervised Learning Representation,CA-SSLR),这是一个可广泛应用于各类语音处理任务的通用条件化模型。与针对下游模型进行优化的标准微调方法不同,CA-SSLR整合来自较早层的语言和说话人嵌入,使SSL模型能够感知当前的语言和说话人上下文。该方法减少了对输入音频特征的依赖,同时保持基础SSLR的完整性。CA-SSLR提升了模型能力,并通过最少的任务特定调整展示了其在未见任务上的通用性。我们的方法采用线性调制来动态调整内部表示,从而在不显著改变原始模型行为的情况下实现细粒度适应性。实验表明,CA-SSLR减少了可训练参数数量,缓解了过拟合,并在资源不足和未见任务中表现优异。具体而言,CA-SSLR实现了LID错误10%的相对降低,在ML-SUPERB基准上ASR CER改善37%,并在VoxCeleb-1上使SV EER降低27%,证明了其有效性。

英文摘要:

We introduce Condition-Aware Self-Supervised Learning Representation (CA-SSLR), a generalist conditioning model broadly applicable to various speech-processing tasks. Compared to standard fine-tuning methods that optimize for downstream models, CA-SSLR integrates language and speaker embeddings from earlier layers, making the SSL model aware of the current language and speaker context. This approach reduces the reliance on input audio features while preserving the integrity of the base SSLR. CA-SSLR improves the model's capabilities and demonstrates its generality on unseen tasks with minimal task-specific tuning. Our method employs linear modulation to dynamically adjust internal representations, enabling fine-grained adaptability without significantly altering the original model behavior. Experiments show that CA-SSLR reduces the number of trainable parameters, mitigates overfitting, and excels in under-resourced and unseen tasks. Specifically, CA-SSLR achieves a 10% relative reduction in LID errors, a 37% improvement in ASR CER on the ML-SUPERB benchmark, and a 27% decrease in SV EER on VoxCeleb-1, demonstrating its effectiveness.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑