A study in Vietnamese statistical parametric speech synthesis base on HMM

Son Thanh Phan, Thang tat Vu, Cuong Tu Duong, Mai Chi Luong

2013 · 11 citations · 8 references

Concepts

Abstract

This article describes an approach in Vietnamese speech synthesis, using statistical parameters speech synthesis system based on hidden Markov models (HMMs), that has grown in popularity over the last few years. Spectral, pitch, tone, and phone duration are simultaneously modeled in HMMs and their parameter distributions are clustered independently by using decision tree-based context clustering algorithms. In this system, statistical modeling is applied to learn distributions of context-dependent acoustic vectors extracted from speech signals, each vector containing a suitable parametric representation of one speech frame and Vietnamese phonetic rules to synthesize speech. Several contextual factors such as tone types, syllables, words, phrases, and utterances were determined and are taken into account to generate the spectrum, pitch, and state duration. The resulting system yields significant correctness for a tonal language, and a fair reproduction of the prosody. Key words: Vietnamese speech synthesis, context dependent, HMM-based, statistical parametric speech synthesis. 1.

References

8