Automatic syllabification using hierarchical hidden Markov models

Author

Nel, Pieter ; du Preez, Johan

Author_Institution

Dept. of Electr. & Electron. Eng., Stellenbosch Univ., South Africa

Volume

1

fYear

2003

fDate

6-10 April 2003

Abstract

This paper presents a purely statistical method for the automatic syllabification of speech. A hierarchical HMM structure is used to implement a purely acoustical model based on the phonotactic constraints found in the English language. A well-defined DTW distance measure is presented for measuring and reporting syllabification results. We achieve a token error rate of 20.3 % with a 42 ms average boundary error on a relatively large set of data. This compares well with previous knowledge- and statistically based methods.

Keywords

error statistics; hidden Markov models; speech processing; speech recognition; statistical analysis; DTW distance measure; English language; acoustical model; automatic syllabification; average boundary error; hidden Markov models; hierarchical HMM; phonotactic constraints; speech syllabification; statistical method; token error rate; Acoustic measurements; Acoustical engineering; Africa; Digital signal processing; Error analysis; Hidden Markov models; Humans; Natural languages; Speech; Statistical analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03). 2003 IEEE International Conference on

ISSN

1520-6149

Print_ISBN

0-7803-7663-3

Type

conf

DOI

10.1109/ICASSP.2003.1198894

Filename

1198894