Syntax Rectification of Speech Using Neural Machine Translation

Speech Recognition is the identification of speech and conversion of the audio signals into textual data. The process of Speech Recognition has been accelerated with advancements in computing and progress in the development of Deep Learning. The mathematics behind Speech Recognition stems from Hidden Markov Models and has evolved into its more advanced form – End to End Speech Recognition. The aim of this paper is to focus on the syntax correction of speech in English language which will make use of Sequence to Sequence Models. The paper will delineate three major domains of grammatical correction of speech which will be achieved through real time interaction with the user.