Indian sign language recognition using convolution neural network

Department of Information Technology, GRIET, India

^* Corresponding author: ledalla.sukanya@gmail.com

Abstract

The goal of the project is to create a machine learning model that can classify the numerous hand motions used in sign language fingerspelling. Communication with deaf and dumb persons is frequently difficult. A variety of hand, finger, and arm motions that assist the deaf and hard of hearing in communicating with others and vice versa. Classification machine learning algorithms are taught on a set of image data in this userindependent model, and testing is done on a completely other set of data. For some people with particular needs, sign language is their only means of communicating their thoughts and feelings. It enables individuals to understand the world around them by visual descriptions and hence contribute to society. As a result, our model aids us in solving the problem more broadly. By watching the user’s hand gestures, this transforms sign language to regular words.