Title :
Mel filter-like admissible wavelet packet structure for speech recognition
Author :
Farooq, O. ; Datta, S.
Author_Institution :
Dept. of Electron. & Electr. Eng., Loughborough Univ. of Technol., UK
fDate :
7/1/2001 12:00:00 AM
Abstract :
A new filter structure using admissible wavelet packet analysis is presented. These filters have the advantage of having frequency bands spacing similar to the Mel scale. Further wavelet packet transform´s multiresolution capabilities are used to derive new sets of features, which are found to be superior to Mel frequency cepstral coefficients (MFCC) in unvoiced phoneme classification problems.
Keywords :
cepstral analysis; feature extraction; filtering theory; pattern classification; speech recognition; wavelet transforms; Mel filter; Mel frequency cepstral coefficients; Mel scale; admissible wavelet packet analysis; conjugate mirror filters; multiresolution capabilities; speech recognition; unvoiced phoneme classification; Binary trees; Cepstral analysis; Discrete wavelet transforms; Filters; Fourier transforms; Mel frequency cepstral coefficient; Signal resolution; Speech recognition; Wavelet analysis; Wavelet packets;
Journal_Title :
Signal Processing Letters, IEEE
