An Efficient Hand Gesture Recognition System Using Deep Learning

Most of the people perform various tasks by using a computer keyboard/mouse leading to repetitive wrist and hand motions, resulting in Carpal Tunnel Syndrome. This paper is geared towards developing a computer management system using hand gestures accomplishing virtual keyboard/mouse operations/commands to effectively eliminate the Carpel Tunnel Syndrome. Gesture Recognition provides an accurate estimation of hand gestures using deep learning algorithm. The complexity of hand structure in obtaining gestures and the rapidness of the movements of the hand or fingers are the problems of tracking algorithms. Thus, deep learning provides a rapid and precise estimate of hand gestures using Convolutional Neural Network (CNN) algorithm. This paper uses articulated CNN algorithm capturing possible gestures, accomplishing various keyboard/mouse operations/commands, thereby avoiding the syndrome. Compared to the conventional algorithm, the proposed work produces high accuracy, a good estimation of hand gestures and cost-effective.