Speech recognition has wider application in now a day, before speech recognition there were a messaging through text which takes a lots of time and efforts and it is a interesting way of communication than text. Therefore people attracted more towards speech than text. Speech recognition is added with new features day by day by using different technologies. Speech recognition is used for lots of category of people such as: handicap, illiterate or people with less knowledge of languages, so that human can enjoy today’s technologies. This paper is based on speech recognition system developed on android platform. System which we have developed is go through different features such as: text-to-speech conversion and vice versa, contact selection by voice and module selection is also voice based, multiple contact selection, frequent message, favorite contacts, notifications etc. As we studied different papers based on this application were having a drawback and fewer features, so as with leads of technologies it is possible to add more features to make it more interesting. We developed this application by using Google API and various algorithms and a dictionary of sentence. For making it voice based we used HMM (hidden markov model).

Keywords : Speech-To-Text, Text-To-Speech Converter (Both Side), HMM (Hidden markov model), frequent contacts, favorite contact, multiple contacts.