A Two-Way Integrated Communication System for the Deaf and Mute

Abstract

A non-disabled person finds it onerous to learn sign language to communicate with the deaf-mutes. This paper discusses a two-way communication system that bridges this communication gap. The proposed system is a portable standalone embedded system that acts as a real-time sign language translator between the deaf-mutes and non-mutes. A fingerspelling data set of Indian sign language that includes some commonly used words were created for training and validating the model. The system uses computer vision techniques and customized speech translator APIs, enabling two-way communication between the deaf-mutes and non-mutes.