Constructing candidate word lists using acoustically similar word groups

Author

Bahl, L.R. ; de Souza, P.V. ; Gopalakrishnan, P.S. ; Kanevsky, D.S. ; Nahamoo, D.

Author_Institution

IBM Thomas J. Watson Res. Center, Yorktown Heights, NY, USA

Volume

40

Issue

11

fYear

1992

fDate

11/1/1992 12:00:00 AM

Firstpage

2814

Lastpage

2816

Abstract

A method for identifying a set of candidate words that matches well with a given utterance is discussed. The method uses precomputed groups of acoustically similar words. Given a vocabulary, for each word a short list of words that are acoustically similar to it is computed. These groups are constructed using training utterances recorded by several speakers and used in an isolated utterance recognition system to construct candidate word lists during recognition. Experiments show that the correct word appears on the short list of candidate words constructed during recognition using this scheme over 98% of the time

Keywords

speech recognition; acoustically similar word groups; candidate word lists; isolated utterance recognition system; speech recognition; training utterances; vocabulary; Acoustic measurements; Hardware; Hidden Markov models; Impedance matching; Loudspeakers; Power system modeling; Real time systems; Signal processing algorithms; Speech recognition; Vocabulary;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.165669

Filename

165669