摘要
针对单通道语音信号盲分离的问题,结合盲源分离和经验模式分解的优点,提出了一种基于经验模式分解的单通道语音信号源数估计和盲源分离方法。对语音混合信号进行经验模式分解,利用贝叶斯算法估计语音源数目,根据源信号数目重组多通道语音混合信号,并采用独立分量分析实现语音信号的盲分离。仿真实验表明,使用此法能有效地估计通道语音信号源数和分离盲源。
In view of the problem of the single channel speech signal blinding separation, using the advantages of the blinding source separation and the empirical mode decomposition, this article proposes a method for estimating the number of the single channel speech signal source and separating the blinding source. It adopts the empirical mode to disintegrating the speech signals, using the Baye algorithm to estimate the speech source number, reorgani- zing the muhichannel pronunciation signal according to the source signal number, and realizing the blinding separation by the method of the isolated component analysis. The simulation experiments show that the proposed method can estimate the number of the single channel speech signal source and separate the blinding source effectively.
出处
《电子科技》
2012年第7期66-68,75,共4页
Electronic Science and Technology
基金
广西研究生教育创新计划基金资助项目(2010105950804M35)
作者简介
颜学龙（1962-），男，硕士，教授，硕士生导师。研究方向：测试计量技术及仪器，信息集成及边界扫描。
赵志强（1985-），男，硕士研究生。研究方向：计算机辅助测试，信号检测及处理。
