Improvements in the stochastic segment model for Phoneme recognition

Open Access

1 January 1989

proceedings article
Published by Association for Computational Linguistics (ACL)

p. 332-338
https://doi.org/10.3115/1075434.1075491

Abstract

The heart of a speech recognition system is the acoustic model of sub-word units (e.g., phonemes). In this work we discuss refinements of the stochastic segment model, an alternative to hidden Markov models for representation of the acoustic variability of phonemes. We concentrate on mechanisms for better modelling time correlation of features across an entire segment. Results are presented for speaker-independent phoneme classification in continuous speech based on the TIMIT database.

Keywords

PHONEME RECOGNITION
HIDDEN MARKOV MODEL
SPEECH RECOGNITION SYSTEM
TIMIT DATABASE
ACOUSTIC VARIABILITY
STOCHASTIC SEGMENT MODEL
ENTIRE SEGMENT
SPEAKER-INDEPENDENT PHONEME CLASSIFICATION
CONTINUOUS SPEECH
BETTER MODELLING TIME CORRELATION
ACOUSTIC MODEL
SPEECH RECOGNITION

This publication has 0 references indexed in Scilit: