Eye Gaze Direction Estimation using Convolution Neural Networks

Currently, eye gaze direction estimation is one of the most pressing problems of computer vision. Gazetracking and eyetracking systems based on machine learning algorithms are used to solve a wide range of tasks. These systems find their application in such areas as marketing, medicine, psychology, human-machine interaction, etc. The subject of this study is the development of client software that tracks the user's eye gaze direction on a monitor using a convolutional neural network. In the course of the work, modern approaches to solving the problem of determining the direction of the gaze were described, a gazetracking application was implemented that runs in the browser, and comparisons were made of the results of this application with the existing one. Based on the experiments, conclusions were made about the program.