A novel technique of sign language recognition system using machine learning for differently abled person

Abstract

Sign language is a language used by deaf and dumb people to communicate through hand gestures or facial expressions combined with non-manual elements. Various automotive tools and software have been developed by many developers, but they require hardware and an Internet connection, which adds to the cost of the software. In this paper, the presented software captures the hand gesture, and with the help of various machine learning optimization algorithms such as Stochastic Gradient Descent (SGD) and Adam (optimizer), the accuracy will be determined to give predictive value. With the help of computers, it could be a new way of learning for deaf and dumb people. During such a pandemic, various group learning apps and software have been developed for the purpose of conducting online classes, but they are useful for ordinary people. Using this technique will help deaf and dumb people with online learning.