 |
International Workshop on Hands-Free Speech Communication (HSC2001)April 9-11, 2001 |
 |
|
International Workshop on Hands-Free Speech Communication (HSC2001)April 9-11, 2001 |
|
A new MINT-based blind separation and deconvolution technique is developed to provide an alternative means of improving the signal-to-noise ratio and reducing reverberation in speech signals. The multiple-input/output inverse-filtering theorem (MINT) [1] is used to compute the stable and accurate multi-channel inverse filters of room impulse responses that may have non-minimum phases. Because the conventional MINT method uses the room impulse responses to calculate the inverse filters, it cannot recover speech signals in practice, where the room impulse responses are unknown in advance. Our method blindly estimates the inverse by computing the correlation matrix between input signals that can be observed, instead of the room impulse responses. The performance of the proposed method is demonstrated using real impulse responses.
Bibliographic reference. Furuya, Ken'ichi (2001): "Noise reduction and dereverberation using correlation matrix based on the multiple-input/output inverse-filtering theorem (MINT)", In HSC2001, 59-62.