Real Time Machine Translation System between Indian Languages

Abstract

Language understanding is one of those perpetual challenges which has dogged research from many decades. As a means of connecting with others, but communication is not limited to a single language. India contains around 121 different languages. As a result, there is a linguistic barrier. Natural Language Processing is the process of developing a communicational interface between machines and humans. A model that turns the supplied source language text (Marathi) into the desired language text (Gujarati) using Text to Text Translation has been developed. Because LSTMs allow us with a vast variety of parameters like as learning rates, input and output biases, and there is no need for existence, this model is constructed by adding multilingual features to the LSTM Model based on deep learning Principles. The goal of the proposed is to create a deep learning-based translation system. The suggested system uses a variety of models to perform text-to-text translation, after which the input text will be translated into the target language (text).