Identifying Indian Sign Languages Using Decentralized Deep Learning

Abstract

Among the main challenges that persons with disabilities encounter on a regular schedule is interaction. The advancement of recognizing signing communicative patterns and techniques made possible by modern technological advancements has significantly contributed to the resolution of this issue. Systems for the identification of hand signals utilizing human intelligence have shown excellent reliability. However, the modeling learning algorithm might take a lot of time because of the vast amount of information needed. We suggest an expedited Indian Signature Identification System that makes use of dispersed intelligence to address this issue. This same concept uses Convolutional Neural Network (CNN) for identifying indications and switching them to English speech. The modeling proposed technique has been distributed over Temporal Unit Operations and Graphics Processors Elements.