Silent speech generation using Brain Machine Interface

Silent speech generation is an intelligent idea that can possibly assist physically challenged people who cannot convey their information as an acoustic signal. Silent speech is generated by predicting the intended speech information which occurs as a result of neural activity involved in the process of speech production. The acquired speech is synthesized and given as a feedback to the user acoustically with the delay of 50ms. This paper briefly elucidate the process of acquiring neural signal, preprocessing and feature extracting for the production of speech signal by means of Brain Machine Interface.