Traditional methods for eye gaze tracking require special cameras and infrared light sources. Such pieces of equipment are usually expensive, and this has limited where the technology can be used. With the advancement of computer vision and neural networks, there has been a lot of interest in performing eye gaze tracking from video images recorded by the built-in cameras of mobile devices, such as smartphones and tablets. This project re-implements GazeCapture, which is a method for eye gaze prediction from a single image. The face and eye regions are cropped and passed into a convolutional neural network. Thereafter, the network predicts the coordinates of the point of the gaze.