| 基于ICA的重叠语音基频提取和语音增强 |
| |
| 引用本文： | 王泽,朱贻盛,王自明,张红煊.基于ICA的重叠语音基频提取和语音增强[J].北京生物医学工程,2001,20(4):241-245,278. |
| |
| 作者姓名： | 王泽 朱贻盛 王自明 张红煊 |
| |
| 作者单位： | 上海交通大学生物医学工程系,200030 |
| |
| 基金项目： | 国家自然科学基金;69871019; |
| |
| 摘 要： | 语音信号是一种特征时变信号，基音频率提取和语音增强是两种常见语音处理要求，独立分量分析（ICA）是一种盲信号处理方法，目的 在于将混合在观察信号中的相互独立的源信号分离出来，ICA在很多领域都有广泛的应用，在语音信号上最为成功，本文将ICA用来撮重叠语音信号中的基音频率，提出了一个基于ICA的混叠语音基频提取系统，成功地解决了基频相互接近时无法提取混合信号基频的问题，并将它应用到有音乐背景下的语音增强方面，均取得了较好的结果。
|
| 关 键 词： | 独立分量分析 语音基频提取 语音增强 语音信号 生物医学工程 |
| 文章编号： | 1002-3208(2001)04-0241-05 |
Independent Component Analysis and its Application to Overlapping Speech Pitch Extraction and Speech Enhancement |
| |
| WANG Ze,ZHU Yisheng,WANG Ziming,ZHANG Hongxuan.Independent Component Analysis and its Application to Overlapping Speech Pitch Extraction and Speech Enhancement[J].Beijing Biomedical Engineering,2001,20(4):241-245,278. |
| |
| Authors: | WANG Ze ZHU Yisheng WANG Ziming ZHANG Hongxuan |
| |
| Institution: | WANG Ze,ZHU Yisheng,WANG Ziming,ZHANG Hongxuan Department of Biomedical Engineering,SJTU 200030 |
| |
| Abstract: | Independent componet analysis(ICA) is a new blind source separating method, with purpose of separating independent sources from mixed signals of them. ICA has been applied in a wide range of signal processing fields. In this paper, ICA is used to do pitch extraction and speech enhancement for overlapping speeches. Good results were obtained in both two aspects. Even in the case when the primary frequencies, contained in overlapping speeches, are very close to each other, the problem was resolved successfully. |
| |
| Keywords: | Independent componet analysis(ICA) Pitch extraction of speech signals |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
