arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.18802cs.SD
收藏

多通道语音增强用于鸡尾酒派对语音情感识别

Multi-Channel Speech Enhancement for Cocktail Party Speech Emotion Recognition

The Chinese University of Hong Kong, Hong Kong SAR, China(香港中文大学);National Research Council Canada(加拿大国家研究委员会);Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所)

Youjun Chen, Guinan Li, Mengzhe Geng, Xurong Xie, Shujie Hu, Huimeng Wang, Haoning Xu, Chengxi Deng, Jiajun Deng, Zhaoqing Li, Mingyu Cui, Xunying Liu

AI总结:

本文提出多通道语音增强方法,用于提升鸡尾酒派对场景中语音情感识别的性能,通过去混响和分离前端结合深度学习模型,实现比单通道方法更高的准确率和F1值。

摘要:

本文强调了多通道语音增强(MCSE)在鸡尾酒派对场景中语音情感识别(ER)中的关键重要性。采用一种多通道语音去混响和分离前端,整合了DNN-WPE和基于掩码的MVDR,用于从混合语音中提取目标说话人的语音,然后将其输入到使用HuBERT和ViT的语音和视觉特征的下游ER后端。实验使用由IEMOCAP和MSP-FACE数据集构建的混合语音数据,表明MCSE输出在一致性上优于由Conformer基于的度量GANs和WavLM SSL特征(可选SE-ER双任务微调)生成的单通道语音表示。在上述单通道基线上,统计显著的加权、无加权准确率和F1测量值分别提高了9.5%、8.5%和9.1%(绝对值为17.1%、14.7%和16.0%的相对值)。此外,IEMOCAP训练的MCSE前端的泛化能力也显示为在零样本应用到外域MSP-FACE数据时。

英文摘要:

This paper highlights the critical importance of multi-channel speech enhancement (MCSE) for speech emotion recognition (ER) in cocktail party scenarios. A multi-channel speech dereverberation and separation front-end integrating DNN-WPE and mask-based MVDR is used to extract the target speaker's speech from the mixture speech, before being fed into the downstream ER back-end using HuBERT- and ViT-based speech and visual features. Experiments on mixture speech constructed using the IEMOCAP and MSP-FACE datasets suggest the MCSE output consistently outperforms domain fine-tuned single-channel speech representations produced by: a) Conformer-based metric GANs; and b) WavLM SSL features with optional SE-ER dual task fine-tuning. Statistically significant increases in weighted, unweighted accuracy and F1 measures by up to 9.5%, 8.5% and 9.1% absolute (17.1%, 14.7% and 16.0% relative) are obtained over the above single-channel baselines. The generalization of IEMOCAP trained MCSE front-ends are also shown when being zero-shot applied to out-of-domain MSP-FACE data.

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。

↑