A Design and Implementation of Hangeul TTS Using Pronouncing Dictionary

Continuous efforts have been doing for simpler and easier interface for human with rapid development of computer technology and its wider use. At present, various voice synthesizing systems and voice recognition systems are commercially available as the result of such an efforts to implements voice synthesizing system and recognition system upon conclusion voice to be the most suitable interface for human and the computer interaction. Among them, voice synthesizing systems have been developed in the form of TTS (Text-to-Speech) domestically, and provide users more convenient interface medium. In this paper, we implement Korean language TTS system with much more improved naturalness in synthesized voice based on MBROLA, an Open source project, which reads Hangeul text file, and selects Pronouncing syllable by searching pronouncing dictionary DB, then extracts rhythm and length information in a series of intonation rule.