使用matla和VC混合编程实现的语音识别-speech recognization

语音信号实时采集，并用独立分量将混合信号分离-Real-time voice signal acquisition, and independent component of mixed-signal separation

基于最大信噪比的盲源分离算法（附参考文章），能够分离线性混合的超高斯和亚高斯源信号, 并且可以有效地分离语音信号。-(With reference to the article), can be separated and super-Gaussian linear mixed sub-Gaussian source signal to noise ratio based on the largest blind source separation algorithm, and can effe