摘要
提出利用核独立分量分析来分离混合语音信号的盲信号处理方法.介绍了基于核函数ICA的原理和基本算法,然后利用该方法分离混合语音信号.实验结果表明:利用基于核独立分量分析的方法可以得到较为理想的波形.
This paper provides a new application method to separate mixed speech signal using Kernel Independent component.Independent component analysis which has developed rapidly during the last few years is one method of the Blind signal processing.Kernel independent component analysis（KICA）＇s principle and algorithm are introduced,and then the KICA is used to separate the mixed signals.The result of experiment shows that with KICA desirable waveform can be obtained.
出处
《重庆工学院学报（自然科学版）》
2009年第11期113-116,122,共5页
Journal of Chongqing Institute of Technology
关键词
独立分量分析
核空间
声音信号
语音分离
independent component analysis Kernel space sound signal speech separation
作者简介
作者简介：段承璋（1986-），男，河南人，硕士研究生，主要从事独立分量分析和小波变换方面的研究．
