A decoding strategy using graph partitioning for continuous-speech recognition

Author

Deller, J.R., Jr. ; Chiu, C.C. ; Yang, Y.P.

Author_Institution

Dept. of Electr. Eng., Michigan State Univ., East Lansing, MI, USA

Volume

3

fYear

1994

fDate

30 May-2 Jun 1994

Firstpage

257

Abstract

Partitioned search (PS) of a language model results in dramatic complexity reduction, and improved recognition rates and noise robustness. PS is applied to recognition of continuous-speech taken from the TIMIT database. Comparisons of the results using the following approaches for recognizing continuous speech are made; a quiet environment, and when the speech is corrupted by intermittent noise: 1. PS, 2. left-to-right search using random selection of as many vertices as in the PS cases for evaluation, and 3. conventional left-to-right stack search with pruning

Keywords

decoding; graph theory; speech recognition; TIMIT database; complexity reduction; continuous-speech recognition; decoding strategy; graph partitioning; intermittent noise; language model; left-to-right search; left-to-right stack search; noise robustness; partitioned search; pruning; quiet environment; recognition rates; Decoding; Laboratories; Natural languages; Noise reduction; Noise robustness; Partitioning algorithms; Speech enhancement; Speech processing; Speech recognition; Working environment noise;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1994. ISCAS '94., 1994 IEEE International Symposium on

Conference_Location

London

Print_ISBN

0-7803-1915-X

Type

conf

DOI

10.1109/ISCAS.1994.409157

Filename

409157