Display control based on eye gaze estimation

Detecting and tracking eye gaze is an active research area and significant progress has been made in this area in the past decades. However, challenge remains due to the differences between individual's eyes, variability in light condition, scale and occlusion. Works on eye location and eye movements have a large number of applications and is an important part of biometrics, human-computer interaction and face detection. Based on the current works and applications on eye gaze detection, the author notice, however, applying eye gaze information on human-computer interaction still lacks of enough work on it. This paper proposes a novel pipeline of employing eye gaze information for display control based on the video captured by integrated camera. The proposed pipeline shows that, despite the low quality of the video and light condition, eye gaze can still be estimated and display of the screen can be controlled accordingly.