摘要
独立分量分析(ICA)作为一种有效的盲源分离技术已成为信号处理领域的热点,但其收敛速度较慢。为此文章简要介绍了有关独立分量分析的基本理论和算法,重点研究了快速定点(FastICA)算法,利用该算法有效地解决了噪声在语音信号中的分离问题。在采集了4个声音信号后,将4个原始信号进行混叠,使用FastICA方法对混叠信号进行分离,将分离结果与原始信号波形进行比对,结果说明该算法具有良好的分离效果。
The independent component analysis, as a method widely used in blind source sepamtion, is a hotspot in signal proeessing, but it converges slowly. The basic prineipe of ICA is discussed in this paper. The FastICA algorithm is studied, since it can effectively solve the problem of the noise in the speech signal. Four actual speech signals are factitiously mixed, and then the mixed signal is separated by using the conventional FastICA. The separated results show that the FastICA has good separation efficiency.
出处
《微型电脑应用》
2017年第1期39-41,共3页
Microcomputer Applications
基金
渭南市科研发展计划项目(2015KYJ-2-6)
渭南师范学院理工类科研项目(16YKS010)
陕西省2017年军民融合研究基金项目(17JMR26)
关键词
语音信号
盲源分离
独立分量分析
快速定点算法
Voice signal
Blind source separation
Independent component analysis
FastICA