Frequency and wavelet filtering for robust speech recognition

Authors:
Murat Deviren;Khalid Daoudi
Affiliations:
INRIA-LORIA, Villers les Nancy, France;INRIA-LORIA, Villers les Nancy, France
Venue:
ICANN/ICONIP'03 Proceedings of the 2003 joint international conference on Artificial neural networks and neural information processing
Year:
2003

Citing 2
Cited 0

Time and frequency filtering of filter-bank energies for robust HMM speech recognition

Speech Communication - Special issue on noise robust ASR
A Wavelet Tour of Signal Processing, Third Edition: The Sparse Way

A Wavelet Tour of Signal Processing, Third Edition: The Sparse Way

Quantified Score

Hi-index	0.00

Visualization

Abstract

Mel-frequency cepstral coefficients (MFCC) are the most widely used features in current speech recognition systems. However, they have a poor physical interpretation and they do not lie in the frequency domain. Frequency filtering (FF) is a technique that has been recently developed to design frequency-localized speech features that perform similar to MFCC in terms of recognition performances. Motivated by our desire to build time-frequency speech models, we wanted to use the FF technique as front-end. However, when evaluating FF on the Aurora-3 database we found some discrepancies in the highly mismatch case. This led us to put FF in another perspective: the wavelet transform. By doing so, we were able to explain the discrepancies and to achieve significant improvements in recognition in the highly mismatch case.