A Device for Automatic Conversion of Speech to Text and Braille for Visually and Hearing Impaired Persons

This paper describes the implementation of a prototype device for individuals dealing with both visual and hearing impairment to communicate. This is carried out with the help of a speech input system and a braille decoder. The input is processed through a machine learning model which recognizes the spoken words, convert spoken words to text, capitalize each letter in the word and send the respective ASCII code to a custom designed decoder through the Raspberry pi 4B module. The decoder is designed using logic gates which are further implemented using n-type and p-type MOSFETs. The proposed device provides an output in the form of braille codes corresponding to each letter which have been implemented using LEDs.