ASSAMESE TRANSLATION SYSTEM

Abstract

Automatic Speech Recognition is the method in which speech signals are translated into the related sequence of characters into text (word). The research on speech recognition has been done for many years ago. After a lot of research work we got the idea of RNN (Recurrent Neural Network) algorithm is more effective and helpful algorithm for speech recognition. The Recurrent Neural Network is the type of Neural Network techniques in which we observe the difference of alphabet. The application or purpose of the RNN is to make people comfortable for the use of Hindi, English and Assamese languages. With the help of this research, people can easily understand languages and get its insight. Languages pose different challenges for speech recognition. This paper discusses about the use of RNN and Statistical Machine Translation algorithms that can be used for translation systems.