Lexical access for speech understanding using minimum message length encoding

Authors:
Ian Thomas;Ingrid Zukerman;Jonathan Oliver;David Albrecht;Bhavani Raskutti
Affiliations:
Department of Computer Science, Monash University, Clayton, Victoria, Australia;Department of Computer Science, Monash University, Clayton, Victoria, Australia;Department of Computer Science, Monash University, Clayton, Victoria, Australia;Department of Computer Science, Monash University, Clayton, Victoria, Australia;Artificial Intelligence Section, Telstra Research Laboratories, Clayton, Victoria, Australia
Venue:
UAI'97 Proceedings of the Thirteenth conference on Uncertainty in artificial intelligence
Year:
1997

Citing 4
Cited 1

Strategies for lexical access to very large vocabularies

Speech Communication - Word Recognition in Large Vocabularies
Fundamentals of speech recognition

Fundamentals of speech recognition
Lexical Access using Minimum Message Length Encoding

PRICAI '96 Proceedings of the 4th Pacific Rim International Conference on Artificial Intelligence: Topics in Artificial Intelligence
The zero-frequency problem: estimating the probabilities of novel events in adaptive text compression

IEEE Transactions on Information Theory

A minimum message length approach for argument interpretation

SIGDIAL '02 Proceedings of the 3rd SIGdial workshop on Discourse and dialogue - Volume 2

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Lexical Access Problem consists of determining the intended sequence of words corresponding to an input sequence of phonemes (basic speech sounds) that come from a low-level phoneme recognizer. In this paper we present an information-theoretic approach based on the Minimum Message Length Criterion for solving the Lexical Access Problem. We model sentences using phoneme realizations seen in training, and word and part-of-speech information obtained from text corpora. We show results on multiple-speaker, continuous, read speech and discuss a heuristic using equivalence classes of similar sounding words which speeds up the recognition process without significant deterioration in recognition accuracy.