Bangla text to speech conversion: A syllabic unit selection approach

Text to speech conversion, abbreviated as TTS, is an artificial voice tool for text documents that converts written text into synthesized sounds. There are quite a number of methods of synthesizing speech for TTS and all of them have their own significant issues like complicated physical model, nonintelligibility, limited usability etc. The best one available is Unit Selection Synthesis, which provides with optimal naturalness and efficiency based on which unit is usedfor example, phoneme, diphone, syllable etc. This paper describes a syllabic method of unit selection synthesis for converting text into speech for Bangla language, the sixth most spoken language in the world. The development process of text collection, text analysis and speech synthesis is also described in this paper.