A comparative analysis of neural network models for hand gesture recognition methods

Abstract

In recent years, gesture recognition methods have undergone major changes. Because the very demand for it has reached a different level. Humans have increasingly begun using various areas of human activity. The purpose of gesture recognition is to record gestures formed in a certain way and then tracked by a device such as a camera. Hand gestures can be used as a form of communication within many different applications. It can be used by people with various disabilities, including those with hearing and speech impairments, to communicate and interact socially with others. Our research demonstrates various methods for implementing hand gesture recognition based on Hidden Markov Model (HMM), Convolutional Neural Network (CNN), Diffractive Deep Neural Network (D2NN) and other neural networks. This research reviews previous approaches and results of hand gesture recognition methods, hypotheses, diagrams, as well as a comparative review between various gesture recognition methods are given in this paper.