State of the art review on thai text-to-speech system

Abstract

Text-to-speech system is a system that converts the input text into speech sound. In Thai language, Thai text-to-speech system was developed in many years. There are many problems with Thai text to speech transformations such as Thai language is a Tonal language that differentiates from others languages. There are 4 main components in Thai text-to-speech synthesis system and several problems in each component. In text analysis, Thai is a language which has no punctuation marks to separate word boundaries and word ambiguity. In letter-to-sound, Thai has several sounds within one word called Homographs. In prosody generation, Thai has five tones which generate many different sounds. Finally, in speech synthesis, naturalness of speech has improved significantly. This paper presents the problems and development of recent researches in Thai text-to-speech system such as Characteristics of the Thai Tonal language, Corpus, Issues in Thai text-to-speech, Recent Thai text-tospeech applications, Future work and Conclusions. © 2008 IEEE.