Large-vocabulary speaker-independent continuous speech recognition using HMM

Author

Lee, Kai-Fu ; Hon, Hsiao-Wuen

Author_Institution

Dept. of Comput. Sci., Carnegie-Mellon Univ., Pittsburgh, PA, USA

fYear

1988

fDate

11-14 Apr 1988

Firstpage

123

Abstract

SPHINX, the first large-vocabulary speaker-independent continuous-speech recognizer is described. SPHINX is a hidden-Markov-model (HMM)-based recognizer using multiple codebooks of various LPC-derived features. Two types of HMMs are used in SPHINX: context-independent phone models and function-word-dependent phone models. On a 997-word task using a bigram grammar, SPHINX achieved a word accuracy of 93%. This demonstrates the feasibility of speaker-independent continuous-speech recognition, and the appropriateness of hidden Markov models for such a task

Keywords

Markov processes; encoding; speech recognition; LPC-derived features; SPHINX; bigram grammar; context-independent phone models; function-word-dependent phone models; hidden-Markov-model; large vocabulary speech recognition; multiple codebooks; speaker-independent continuous speech recognition; word accuracy; Autocorrelation; Cepstral analysis; Cepstrum; Computer science; Context modeling; Hidden Markov models; Linear predictive coding; Speech processing; Speech recognition; System testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

Conference_Location

New York, NY

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1988.196527

Filename

196527