Abstract
A human gesture is a non-verbal form of communication and is critical in human-robot interactions. Vision-based gesture recognition methods play a key role to detect hand motion and support such interactions. Hand gesture recognition allows a convenient and usable interface between devices and users. Hand gestures can be used for various fields which makes it be able to be implemented for communication and further. Hand gesture recognition is not only useful for people who are hearing-impaired or disabled but also for the people who experienced a stroke, as they need to communicate with other people using different common essential gestures such as the sign of eating, drink, family and, more. In this paper, an approach for recognizing hand gesture based on Convolutional Neural Network (CNN) is proposed. The developed method is evaluated and compared between training and testing modes based on several metrics such as execution time, accuracy, sensitivity, specificity, positive and negative predictive value, likelihood and root mean square. Results show that testing accuracy is 99% using CNN and is an effective technique in extracting distinct features and classifying data
Keywords – Convolutional Neural Network, Deep Learning, Hand Gesture Recognition, Gesture Recognition


INTRODUCTION

Recently, direct contact is the predominant form of communication between the user and the machine. The communication channel is based on devices such as a mouse, keyboard, remote control, touch screen, and other direct contact methods. Human to human communication is accomplished through more natural and intuitive non-contact methods, for example, sound and physical movements. The flexibility and efficiency of these non-contact communication methods have led many researchers to consider employing them to support human-computer interaction. The gesture is an important non-contact human communication method which forms a substantial part of the human language. Historically, wearable data gloves were regularly used to capture the angles and positions of each joint in the user's gesture. The difficulty and cost of a wearable sensor have restricted the widespread use of such a method. Gesture recognition can be defined as the ability of a computer to understand the gestures and perform certain commands based on those gestures. The main goal of Gesture recognition is to develop a system that can identify and understand specific gestures and communicates information from them [1].
Gesture recognition methods based on the non-contact visual inspection are currently popular. This is due to their low cost and convenience to the user. A hand gesture is an expressive communication method used in healthcare, entertainment and education industry, in addition to assisting users with special needs and the elderly. Hand tracking is vital to perform hand gesture recognition, involves undertaking various computer vision operations including hand segmentation, detection, and tracking.
Sign language uses hand gestures to convey feelings or information within the hearing impairment communication. The main problem is that an ordinary person would easily misunderstand the meaning conveyed. The advancement in AI and computer vision can be adapted to recognize and learn the sign language [2]. The modern systems can help an ordinary person to recognize and understand the sign language. This article presents a method which is related to the recognition of hand gestures using deep learning.
Stroke is a disease that affects arteries leading to and within the brain. Stroke is the fifth leading death cause as well as a cause of disability. A stroke occurs when a blood vessel that carries oxygen and nutrients to the brain is either blocked by a clot or bursts. Certain security measures keep the privacy implemented and protects an important part of the profile. This information has gathered people to demand skilled and capable information. Networks have a medical diagnosis system to allow the users in the expertise and experiences of groups and individual. This project shows that hand gesture is a very beneficial way to convey information and a very rich set of feelings and facts can be interpreted from gestures.
The rest of the paper is structured as follows: The material and method used in this paper found in Section II. A literature review of hand gesture detection techniques and methods used are shown in Section III. Theoretical concept of CNN is provided in Section IV. Section V concentrates on the details of the proposed system's implementation. Section VI describes the discussion and presentation of the results obtained.


Objective
Hand gestures can be used for various fields which makes it be able to be implemented for communication and further. South Indian sign languages recognition is not only useful for people who are hearing-impaired or disabled but also for the people who experienced a stroke, as they need to communicate with other people using different common essential gestures such as the sign of eating, drink, family.