The decipher speech recognition system

Author

Cohen, Moshik ; Murveit, Hy ; Bernstein, Jared ; Price, Patti ; Weintraub, Mitch

Author_Institution

SRI Int., Menlo Park, CA, USA

fYear

1990

fDate

3-6 Apr 1990

Firstpage

77

Abstract

A large-vocabulary, continuous-speech system, called DECIPHER, which is based on a hidden Markov model (HMM) approach and is designed to achieve high word accuracy in a speaker-independent mode is described. The results of a series of experiments that test acoustic and phonological adaptation of the DECIPHER system to the pronunciations of a single speaker in a speaker-dependent task are presented. Estimating the probabilities of alternative pronunciations and speaker-dependent phonology are discussed

Keywords

speech analysis and processing; speech recognition; DECIPHER system; HMM; acoustic adaptation; continuous-speech system; hidden Markov model; high word accuracy; large-vocabulary; phonological adaptation; probabilities of alternative pronunciations; speaker-dependent phonology; speaker-dependent task; speaker-independent mode; speech recognition system; Acoustic testing; Context modeling; Databases; Discrete Fourier transforms; Filters; Hidden Markov models; Loudspeakers; Resource management; Speech recognition; System testing; Testing; Training data; Vocabulary;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Conference on

Conference_Location

Albuquerque, NM

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1990.115541

Filename

115541