A method for continuous speech segmentation using HMM

Author

Nakagawa, Seiichi ; Hashimoto, Yasuhide

Author_Institution

Dept. of Inf. & Comput. Sci., Toyohashi Univ. of Technol., Toyohashi, Japan

fYear

1988

fDate

14-17 Nov 1988

Firstpage

960

Abstract

A statistical method of segmentation using a hidden Markov model (HMM) and a Bayesian classifier is described. The main features of this method are the use of feature parameters which are independent of each category in vowels of consonants, and the use of only one HMM which commonly represents all syllable patterns. The segmentation strategy is to find the optimal HMM sequence. The optimal/best sequence is found by using the O(n) DP matching based on Viterbi algorithm. The concatenated number and boundaries of the best HMM sequence are regarded as the segmentation result. The experimental result on Japanese spoken sentences shows that the rate of segmentation is more than 92% for two male speakers, and the rate is improved to 97.5% by using a duration control mechanism based on a discrete probability distribution

Keywords

Bayes methods; Markov processes; speech recognition; statistical analysis; Bayesian classifier; Japanese; Viterbi algorithm; consonants; continuous speech segmentation; discrete probability distribution; hidden Markov model; optimal/best sequence; speech recognition; statistical method; syllable patterns; vowels; Automatic speech recognition; Concatenated codes; Hidden Markov models; Integrated circuit modeling; MONOS devices; Probability distribution; Speech processing; Speech recognition; Statistical analysis; Viterbi algorithm;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition, 1988., 9th International Conference on

Conference_Location

Rome

Print_ISBN

0-8186-0878-1

Type

conf

DOI

10.1109/ICPR.1988.28414

Filename

28414