An Optimum Approach to Indian Sign Language Recognition using Efficient Convolution Neural Networks

The most important component of every discussion is language, and sign language is the most formal mode of communication for persons who are deaf or mute. Communication, as we all know, is one of the most important components of modern life. The importance of communication is increasing in tandem with our maturation in life. The purpose of this project is to develop a system capable of recognizing all of these signs and serving as a bridge for signers and non-signers to communicate and interpret sign meanings. Technology is rapidly evolving, and now that we have powerful systems, a wealth of data, and a diverse set of sources, we can design a system to assist people with disabilities. We will develop a user-independent framework for automatically recognizing Indian Sign Language in this project, which will allow for feature extraction, recognition, and interpretation of a variety of one-handed, dynamic isolated signals. The two most important and fundamental aspects of SLR research are isolated sign recognition and continuous sign recognition. Hand form, hand position, and so on are the primary hand traits that can be used in sign language. The extra-clung functionality can be implemented in two ways. Either one is tracking-based or it is not. This essay focuses on one such approach that we devised to overcome significant communication barriers. The technique is divided into three stages: pre-processing, texture extraction, and recognition.