语音信号盲分离方法研究  

Method of speech signal blind separation

作　　者：孙小军[1] 

机构地区：[1]平顶山学院,467099

出　　处：《电子测试》2013年第12期25-27,共3页Electronic Test

基　　金：平顶山学院青年基金（编号：PDSU-ONJJ-2013006）

摘　　要：本文采用WVA分布与联合对角化的盲分离算法，估计出源语音信号，实现对混叠信号的盲分离。通过仿真实验，结果表明，本算法具有分离效果好，能有效的将混叠的盲语音信号分离。The blind source separation refers to the process of each original signal only mixed signals from several observation to distill, recovery cannot direct observations of the. In this paper, the Wegener time- frequency distribution and the joint diagonalization of the blind separation method based on speech signal, estimate the source, through the experiment, and the results show that the algorithm can separate blind speech signals effectively.

关 键 词：盲分离 语音信号 盲源信号分离 

分 类 号：TN912.3[电子电信—通信与信息系统]