Indian Sign Language Recognition for Hearing Impaired: A Deep Learning based approach

People with hearing impairment or hearing disability needs sign language for communication. A sign language in America popularly known American Sign Language (ASL), in China known as Chinese Sign Language (CSL) are being used by over 45 million people with hearing impairment to bridge the communication gap with ordinary people. Despite of that, the use of Indian Sign Language (ISL) in India is very primitive or unknown to many ordinary people. Ordinary people having very meagre knowledge over this sign language, makes it difficult to bridge the communication gap with impaired once. So, the system that recognizes sign language can help both impaired people and ordinary people in communicating with each other, not only that it can be very crucial for them in emergency situations. Because of increase increasing growth of hearing impaired this field of research has gained significant importance, moreover its wide range of application including video games, Vedic mathematics, mental and fast calculations etc. has made it more significant. This article presents a Statc Sign language recognition, which aims to detect a specific motion of hands, identify and classify them using useful Transformer Encoder tool Vision Transformers (ViT). Deep learning-based recognition models such as Transformer encoders, Pre-trained CNN models such has ResNet/VGG-16 can be used in implementation of recognition tasks.