Indian Sign Language Character Recognition System

Abstract

Sign language is a vital mode of communication for the deaf community worldwide. However, even to this day, there exists an evident communication gap between signers and non-signers. To address this problem, researchers have proposed automatic sign language recognition systems. Despite extensive research on American Sign Language (ASL), there is limited focus on Indian Sign Language (ISL). The approach in this paper tries to explore the challenges and opportunities involved in developing an automatic sign language recognition system for ISL. The aim was to narrow down the communication gap for the hearing impaired communities and provide a tool for better accessibility. The paper identifies the gesture recognition and skin segmentation challenges while trying to provide an end product capable of recognizing different characters of ISL. Deep learning models were experimented to obtain a good accuracy score while also keeping the process cost and resource efficient.