Automatic segmentation and labeling based on boundary type for mandarin Chinese speech

Author

Huang, Ping-mu ; Sun, Cheng-Li ; Guo, Jun

Author_Institution

Sch. of Inf. Eng., Beijing Univ. of Posts & Telecommun., Beijing

Volume

5

fYear

2008

fDate

12-15 July 2008

Firstpage

2497

Lastpage

2502

Abstract

This paper presents an automatic segmentation and labeling method for mandarin chinese speech synthesis corpus. In order to improve the accuracy of segmentation, two types of HMM models are utilized to produce the INITIAL/FINAL and syllable boundaries. Three feature detection algorithms are applied to boundary refinement for speech boundaries of voiced/unvoiced/silence. Experimental results show the proposed method can improve the performance of the segmentation system significantly.

Keywords

hidden Markov models; natural language processing; speech processing; speech synthesis; Mandarin Chinese speech; Mandarin Chinese speech synthesis corpus; boundary refinement; segmentation system; speech automatic segmentation; speech boundaries; speech labeling; Computer vision; Context modeling; Cybernetics; Detection algorithms; Hidden Markov models; Labeling; Machine learning; Mel frequency cepstral coefficient; Speech synthesis; Sun; HMM; Measurement of the distance between segments; Phoneme boundary-type; Speech segmentation; V/U/S detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2008 International Conference on

Conference_Location

Kunming

Print_ISBN

978-1-4244-2095-7

Electronic_ISBN

978-1-4244-2096-4

Type

conf

DOI

10.1109/ICMLC.2008.4620828

Filename

4620828