This paper proposes a post-filtering estimation scheme for multichannel noise reduction. The proposed method extends and improves the existing Zelinskis and, the most general and prominent, McCowans post-filtering methods that use the auto- and cross-spectral densities of the multichannel input signals to estimate the transfer function of the Wiener post-filter. A major drawback of these two speech enhancement algorithms is that the noise power spectrum at the beamformers output is over-estimated and therefore the derived filters are sub-optimal in the Wiener sense. The proposed method deals with this problem and can be considered as an optimal postfilter that is appropriate for a wide variety of different noise fields. In experiments over real-noise multichannel recordings, the proposed technique is shown to obtain a significant headstart over the other methods in terms of signal-to-noise ratio and speech degradation measures. In addition it is used for ASR experiments where promising preliminary results are presented.
Cite as: Leukimmiatis, S., Dimitriadis, D., Maragos, P. (2006) An optimum microphone array post-filter for speech applications. Proc. Interspeech 2006, paper 1389-Wed3FoP.4, doi: 10.21437/Interspeech.2006-557
@inproceedings{leukimmiatis06_interspeech,
author={Stamatis Leukimmiatis and Dimitrios Dimitriadis and Petros Maragos},
title={{An optimum microphone array post-filter for speech applications}},
year=2006,
booktitle={Proc. Interspeech 2006},
pages={paper 1389-Wed3FoP.4},
doi={10.21437/Interspeech.2006-557}
}