Unsupervised pronunciation grammar generation for non-native speech recognition

Author

Huang, Chien-Lin ; Wu, Chung-Hsien ; Chen, Yi ; Hsu, Chin-Shun ; Lee, Kuei-Ming

Author_Institution

Nat. Cheng Kung Univ., Tainan

fYear

2007

fDate

Oct. 30 2007-Nov. 2 2007

Firstpage

1

Lastpage

4

Abstract

This study presents a novel approach to unsupervised pronunciation grammar generation for non-native speech recognition. Unsupervised pronunciation grammar generation includes pronunciation variation graph construction, stochastic Markov search and grammar selection. Context-dependent relation and phone broad class information are used for variation graph construction. Confidence measure and co-occurrence frequency are used to select the variants of pronunciation grammar for non-native speech modeling. Experiments show that unsupervised pronunciation grammar generation is suitable for the improvement of non-native speech recognition.

Keywords

Markov processes; grammars; speech recognition; context-dependent relation; grammar selection; nonnative speech modeling; nonnative speech recognition; phone broad class information; pronunciation variation graph construction; stochastic Markov search; unsupervised pronunciation grammar generation; variation graph construction; Automatic speech recognition; Computer industry; Computer science; Construction industry; Frequency; Industrial relations; Natural languages; Speech processing; Speech recognition; Stochastic processes;

fLanguage

English

Publisher

ieee

Conference_Titel

TENCON 2007 - 2007 IEEE Region 10 Conference

Conference_Location

Taipei

Print_ISBN

978-1-4244-1272-3

Electronic_ISBN

978-1-4244-1272-3

Type

conf

DOI

10.1109/TENCON.2007.4428886

Filename

4428886