Microphone array speech enhancement based on a generalized post-filter and a novel perceptual filter

The theoretic foundation of traditional microphone array post-filters is the assumption that the noise between sensors is uncorrelated. However, this assumption is inaccurate in real environments since the correlated noise exists. In this paper, a generalized microphone array post-filter is proposed to deal with both the correlated and uncorrelated noise in environments and a novel perceptual filter is proposed to reduce the musical residual noise introduced by the post-filter. Experiments show that the proposed technique produces impressive results in terms of quality measures of the enhanced speech.