04-11-2014, 10:22 AM
Abstracts: Software is available in the market that provides text to speech conversion in English. Gujarati being a native and regional language no support was available. It was a research and development project which required making of ‘Gujarati phonetic engine’ which is capable of reading Gujarati characters and generating respective language specific speech as per the linguistic phonemes. Project deals with selecting the respective phonemes from the file written in Gujarati as input. With the best search technique we decide the most probable phoneme near to the text written in the file. With the respective grammar rules pertaining to the text, the sentence is framed. Any computer system generates sound. The system has a preference for using that sound to create voices for the respective languages. Using system programming we generate the respective sound or we search from the existing recorded sound files we have stored in the system. With due respective to system time of searching and then playing of the sound – the system will speak the text written in any kind of file.