This paper describes a novel application of multiresolution analysis (MRA)
in extracting acoustic features that possess de-noising capability for robu
st speech recognition. The MRA algorithm is used to construct a mel-scaled
wavelet packet filter-bank, from which subband powers are computed as the f
eature parameters for speech recognition. Wiener filtering is applied to a
few selected subbands at some intermediate stages of decomposition. For hig
h-frequency bands, Wiener filters are designed based on a reduced fraction
of the estimated noise power, making the consonant features much more promi
nent and contrastive. The proposed method is evaluated in phone recognition
experiments with the MIT database. In the presence of stationary white noi
se at 10-dB SNR, the de-noised MRA features attain a phone recognition rate
of 32%. There is a noticeable improvement compared with the accuracy of 29
% and 20% attained by the commonly used mel-frequency cepstral coefficients
(MFCC) with and without cepstral mean normalization (CMN), respectively. T
he effectiveness of the MRA features is also verified by the fact that they
exhibit smaller distortion from clean speech. (C) 2001 Acoustical Society
of America.