Developing a hand gesture recognition Web system

With the popularization of different ways to interact with computers, gesture recognition is an area of increasing interest among researchers around the world because of how it enables a user to control an application by performing static hand gestures in front of a camera connected to his machine. In this project, therefore, a Web application that allows users to create their own systems that are capable of recognizing the static gestures that they wanted without the need for a computer with a high processing power was developed, the application also allowed the user to store and load the generated system, avoiding the data loss and provides incentive for sharing of usercreated systems. After the implementation of a Convolutional Neural Network that is based in a MobileNet model, this work was validated by sequences of tests that displayed the feasibility to create systems capable of recognizing gestures with high accuracy.