一种语音信号欠定盲分离的新方法  

A New Method of Underdetermined Blind Separation for Voice Signals Separation for Voice Signals

作　　者：白琳[1] 陈豪[1] 

机构地区：[1]中国空间技术研究院西安分院,西安710000

出　　处：《空间电子技术》2012年第2期16-19,65,共5页Space Electronic Technology

摘　　要：文章提出了一种语音信号欠定盲分离的新方法,引入了一种伪提取矢量的概念,并通过在信号采样点处选取对应的伪提取矢量,以恢复取值占优的源信号采样点的值,来实现欠定盲源分离。通过将该方法与传统的基于线性规划的欠定盲源分离方法分别仿真表明:文章所提出方法由于在信号的各采样点处无需优化,因此大大提高了信号分离的速度,语音信号的分离速度比基于线性规划的方法快数倍。A new method is put forward for accomplishing underdetermined blind signal separation(UBSS) of voice signals in the paper.A new concept of pseudo extraction vectors is introduced.The method accomplishes UBSS by choosing corresponding pseudo extraction vector to recover the signal dominating for each sampling.Some conclusions can be reached by separately simulating the method based on pseudo extraction vectors and linear programming: the method based on pseudo extraction vectors free of optimizing process increases the velocity of separating source signals.The velocity of separation of the method based on pseudo extraction vectors is several-fold of the method based on linear programming for voice signals.

关 键 词：语音信号 欠定盲源分离 稀疏表征 伪提取矢量 

分 类 号：TN91[电子电信—通信与信息系统]