Number Recognition System-based Virtual Sketch with Hand Gestures Using Attentional Convolutional Network

Abstract

Advances in computer technology in the modern era have proliferated. Computer technology has had a significant impact on human life. The relationship between humans and computers, known as Human-Computer Interactions (HCI), can be done non-contact or virtual; one example is Hand Gesture Recognition. However, the development of hand gesture recognition has several challenges, especially in increasing the performance of its glory. Several methods are used to improve the performance of hand gesture recognition, such as the convolutional neural network, which still needs improvement. In this paper, we conducted a number recognition system-based virtual sketch with a hand gesture using an Attentional Convolutional Network with an accuracy value obtained in the non-real-time of 83.9% and real-time accuracy is 90%.