Original scientific paper

Continuous Speech Recognition by a Network of Hidden Markov Models

Nikola Pavešić ; Faculty of Electrical and Computer Engineering, University of Ljubljana, Ljubljana, Slovenia

Abstract

This paper gives an overview of the principles of a system for phoneme based, large vocabulary, continuous speech recognition. In particular, the issue of acoustic-phonetic modelling using Hidden Markov Models is addressed.

Some experiences of Laboratory for artificial perception at University of Ljubljana in attacking some of these problems is presented.

Keywords

Continuous Speech Recognition; Hidden Markov Model; Viterbi Algorithm; Baum-Welch Algorithm; Acoustic-Phonetic Modelling; Language Modelling; Perplexity; Viterbi-Decoding Technique; Beam Search

Hrčak ID:

150427

URI

https://hrcak.srce.hr/150427

Publication date:

30.9.1995.

Visits: 502 *