Speech pattern generation and recognition using a personal computer

Author

Gopalan, K. ; Burridge, B.G.

Author_Institution

Dept. of Eng., Purdue Univ. Calumet, Hammond, IN, USA

fYear

1988

fDate

0-0 1988

Firstpage

268

Lastpage

272

Abstract

A personal-computer-based system to generate and recognize speech patterns of discrete utterances from a limited vocabulary is described. The system implements a time-domain approach based on successive amplitude samples to develop a nonparametric characterization of utterances. Unknown words are recognized using a multiprototype minimum-distance classifier. The method is computationally simple and fast, and yields better word-recognition results than previously reported.<>

Keywords

computerised pattern recognition; microcomputer applications; speech recognition; speech synthesis; discrete utterances; limited vocabulary; multiprototype minimum-distance classifier; nonparametric characterization; pattern generation; pattern recognition; personal-computer-based system; speech patterns; speech recognition; time-domain approach; unknown words; word-recognition; Autocorrelation; Background noise; Character recognition; Microcomputers; Pattern matching; Pattern recognition; Signal generators; Speech recognition; Time domain analysis; Vocabulary;

fLanguage

English

Publisher

ieee

Conference_Titel

System Theory, 1988., Proceedings of the Twentieth Southeastern Symposium on

Conference_Location

Charlotte, NC, USA

ISSN

0094-2898

Print_ISBN

0-8186-0847-1

Type

conf

DOI

10.1109/SSST.1988.17058

Filename

17058