Eye Tracking in Augmented Spaces: A Deep Learning Approach

The use of deep learning for estimating eye gaze in augmented spaces is investigated in this work. There are two primary ways of interacting with augmented spaces. The first involves the use of AR/VR systems; the second involves devices that respond to the user's gaze directly. This domain can overlap with AR/VR environments but is not exclusive to them and contains its own unique set of issues. Deep learning methods for eye tracking that are capable of performing with minimal power consumption are investigated for both problems.