Selection of the best wavelet base for speech signal

Author

Long, Yan ; Gang, Liu ; Jun, Guo

Author_Institution

Sch. of Inf. Eng., Beijing Univ. of Posts & Telecommun., China

fYear

2004

fDate

20-22 Oct. 2004

Firstpage

218

Lastpage

221

Abstract

In this paper a new method is proposed for choosing the best wavelet base for speech signals. The wavelets of Haar, Daubechies, Meyer, biorthogonal, coiflets, and symlets are used for speech signals. The wavelet decomposition is used to decompose speech signals into five levels. The low-frequency part of the signal was reconstructed. By calculating the maximum error and mean error of the reconstructed signal, the results show that the Haar wavelet is not suitable for speech signals and Dmey is the best wavelet base for speech signals.

Keywords

Haar transforms; signal reconstruction; speech processing; wavelet transforms; Daubechies wavelet; Dmey wavelet base; Haar wavelet; Meyer wavelet; biorthogonal wavelets; coiflets; low-frequency signal reconstruction; maximum error; mean error; speech signal; speech signal decomposition; symlets; wavelet decomposition; Continuous wavelet transforms; Discrete wavelet transforms; Filters; Frequency; Hidden Markov models; Postal services; Signal analysis; Speech analysis; Speech processing; Wavelet analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Multimedia, Video and Speech Processing, 2004. Proceedings of 2004 International Symposium on

Print_ISBN

0-7803-8687-6

Type

conf

DOI

10.1109/ISIMP.2004.1434039

Filename

1434039