Gesture recognition sign conversion using deep learning

Abstract

An integral part of human interaction is speech. Human beings talk to each other to share their emotions thoughts and experiences. This however isn't the case for mute people. This paper presents a Gesture recognition sign conversion using deep learning. A sign language is a language used by mute

and hearing-impaired people to communicate. It is a linguistic system based on hand motions, expressed through manual articulations in combination with non-manual elements which facilitates communication through body/facial postures, expressions and a set of gestures. In order to contribute the well being of the affected population, we are motivated to implement a vision¬ based system to prevent their daily life challenges. We propose a methodology for the recognition of hand gestures based on an efficient deep convolution neural network (CNN) architecture. To achieve the robustness performance, the analysis of hand gestures and orientation are applied to urge the training and testing data for the CNN. Our propose model includes object detection and classification phases. The module has single shot detector (SSD) used for hand detection plus a fully connected network utilized to constructively translate the detected signs in to text. In this experiment, we have provided a validation of the proposed method on recognizing human gestures which shows robust results with various hand positions. The experimental results indicate the feasibility and reliability of the proposed system.