结合GSC与相干滤波器的二元麦克风阵列语音增强算法

Speech enhancement method based on the combination of GSC and coherence filter for small microphone arrays

提出一种相干滤波器与广义旁瓣相消器结合(GSC)的二元麦克风阵列语音增强算法.将基于噪声谱估计的单通道相干滤波器作为广义旁瓣相消器的后置滤波器,充分利用阵元间蕴含的信号进行噪音抑制,克服经典结合算法无法使用基于噪声谱估计的相干滤波器的缺点.计算机仿真实验表明,该算法明显优于小阵列广义旁瓣相消算法和基于相位差的算法.

This paper proposed a speech enhancement method which based on the combination of generalized sidelobe canceller（GSC）and coherence filter for small microphone array.The use of single-channel coherence filter based on noise spectrum estimation as a generalized sidelobe canceller post filter ,which takes full advantage of information between array elements to suppress noise and to overcome the shortcomings of classical algorithms which can not use coherence filter that based on the noise spectrum estimation.Simulations tests showed that the proposed algorithm was superior to separate small arrayGSCalgorithms and phase difference algorithm.