USEF-TSE: Universal Speaker Embedding Free Target Speaker Extraction（通用无说话人嵌入目标说话人提取）
USEF-TSE: Universal Speaker Embedding Free Target Speaker Extraction
AI总结：
针对传统目标说话人提取依赖说话人识别模型及嵌入的问题，提出无需说话人嵌入的USEF-TSE框架，以多头交叉注意力提取帧级特征，可适配多种分离模型，在多个标准基准及分布外数据上取得SOTA性能。
摘要：
目标说话人提取旨在从混合语音中分离出特定说话人的声音。传统上，该过程依赖于从参考语音中提取说话人嵌入，这需要用到说话人识别模型。然而，找到合适的说话人识别模型颇具挑战性，且使用目标说话人嵌入作为参考信息对于目标说话人提取任务而言可能并非最优选择。\n本文提出了一种通用无说话人嵌入目标说话人提取（Universal Speaker Embedding-Free Target Speaker Extraction，USEF-TSE）框架，该框架无需依赖说话人嵌入即可运行。USEF-TSE采用多头交叉注意力机制作为帧级目标说话人特征提取器。这种创新方法使主流说话人提取方案能够摆脱对说话人识别模型的依赖，并更好地利用注册语音中可用的信息，包括说话人特征和上下文细节。此外，USEF-TSE可与其他时域或时频域语音分离模型无缝集成，以实现有效的说话人提取。\n实验结果表明，在WSJ0-2mix、WHAM!和WHAMR!数据集（单声道无回声、含噪及含噪混响双说话人语音分离和说话人提取的标准基准）上，本文提出的方法在尺度不变信号失真比（Scale-Invariant Signal-to-Distortion Ratio，SI-SDR）指标上达到了当前最优（state-of-the-art，SOTA）性能。在LibriMix数据集和ICASSP 2023 DNS挑战赛盲测集上的结果表明，该模型在更多样化的分布外数据上表现良好。源代码可访问以下链接获取：https://github.com/ZBang/USEF-TSE。
英文摘要：
Target speaker extraction aims to separate the voice of a specific speaker from mixed speech. Traditionally, this process has relied on extracting a speaker embedding from a reference speech, in which a speaker recognition model is required. However, identifying an appropriate speaker recognition model can be challenging, and using the target speaker embedding as reference information may not be optimal for target speaker extraction tasks. This paper introduces a Universal Speaker Embedding-Free Target Speaker Extraction (USEF-TSE) framework that operates without relying on speaker embeddings. USEF-TSE utilizes a multi-head cross-attention mechanism as a frame-level target speaker feature extractor. This innovative approach allows mainstream speaker extraction solutions to bypass the dependency on speaker recognition models and better leverage the information available in the enrollment speech, including speaker characteristics and contextual details. Additionally, USEF-TSE can seamlessly integrate with other time-domain or time-frequency domain speech separation models to achieve effective speaker extraction. Experimental results show that our proposed method achieves state-of-the-art (SOTA) performance in terms of Scale-Invariant Signal-to-Distortion Ratio (SI-SDR) on the WSJ0-2mix, WHAM!, and WHAMR! datasets, which are standard benchmarks for monaural anechoic, noisy and noisy-reverberant two-speaker speech separation and speaker extraction. The results on the LibriMix and the blind test set of the ICASSP 2023 DNS Challenge demonstrate that the model performs well on more diverse and out-of-domain data. For access to the source code, please visit: https://github.com/ZBang/USEF-TSE.
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。
