Wavelet neural network for classification of transient signals

Author

Tollig, CJA ; Hoffman, AJ

Author_Institution

Sch. of Electr. & Electron. Eng., Potchefstroom Univ. for CHE, South Africa

fYear

1997

fDate

9-10 Sep 1997

Firstpage

161

Lastpage

166

Abstract

A method is presented for adaptively generating wavelet templates for pattern representation. The idea is to form “super-wavelets” that allows the shape of the wavelet to adapt to each presented pattern. The super-wavelets form compact features of the signal which can be dilated or translated to provide scale and position invariant classification. These wavelets will form a bank of filters that can be correlated with the input patterns. The correlation peaks would identify the patterns. We demonstrate the extraction of super-wavelets for speech signals and describe how this feature extraction technique can be employed as part of a classifier

Keywords

adaptive signal processing; feature extraction; feedforward neural nets; pattern classification; signal representation; speech processing; speech recognition; transients; wavelet transforms; adaptive wavelet templates generation; compact features; correlation peaks; dilated signals; feature extraction; filter bank; input patterns; pattern representation; phoneme recognition; position invariant classification; radial basis function; scale invariant classification; signal representation; speech signals; super-wavelets; transient signals classification; translated signals; wavelet neural network; Channel hot electron injection; Continuous wavelet transforms; Feature extraction; Filter bank; Neural networks; Radial basis function networks; Shape; Signal generators; Speech; Time frequency analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications and Signal Processing, 1997. COMSIG '97., Proceedings of the 1997 South African Symposium on

Conference_Location

Grahamstown

Print_ISBN

0-7803-4173-2

Type

conf

DOI

10.1109/COMSIG.1997.630002

Filename

630002