Evaluation of wavelet filters for speech recognition

Author

Kim, Kidae ; Youn, Dae Hee ; Lee, Chulhee

Author_Institution

Dept. of Electr. & Comput. Eng., Yonsei Univ., Seoul, South Korea

Volume

4

fYear

2000

fDate

2000

Firstpage

2891

Abstract

Since wavelet decomposition of signals provides more flexible time-frequency resolutions, it can be utilized as a feature set for speech recognition. The authors explore the possibility of using wavelet decomposition for speech recognition. In particular, they investigate a modified octave structured 5-level filter bank and the HMM (hidden Markov model) is used as a recognizer. We present an analysis of various wavelet filters for speech recognition and compare the results with the conventional features that include LPC and mel-cepstrums

Keywords

filters; hidden Markov models; linear predictive coding; speech recognition; wavelet transforms; HMM; LPC; feature set; flexible time-frequency resolutions; hidden Markov model; mel-cepstrums; modified octave structured 5-level filter bank; signal decomposition; speech recognition; wavelet decomposition; wavelet filter evaluation; Band pass filters; Channel bank filters; Feature extraction; Filter bank; Hidden Markov models; Image coding; Signal resolution; Speech recognition; Time frequency analysis; Wavelet transforms;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man, and Cybernetics, 2000 IEEE International Conference on

Conference_Location

Nashville, TN

ISSN

1062-922X

Print_ISBN

0-7803-6583-6

Type

conf

DOI

10.1109/ICSMC.2000.884438

Filename

884438