Implementation of Subachan: Bengali text to Speech Synthesis Software

Abstract

This paper discusses the design and development of Text-to-Speech for Bengali language. The major modules of our system are Normalization, Phonetic analysis, Prosodic analysis and Wave synthesis. In normalization process, we have used some modules for example token identification, lookup table, expansion rules for analysing a sentence. Using these modules, we can recognize the type of each word clearly and find out the dependency of the words in a sentence. Normalization solves ambiguity problem and increases correctness. In Phonetic analysis, we have used grapheme to phoneme rules for most of the cases but to solve the problem of O-karanto problem we have used a small dictionary containing the pronunciation of couple of words. And finally we have applied concatenation approach on diphone to develop the system. In ideal situation, our system can produce better performance. And with some exceptions, our synthesis system works well in any situation.