In order to better understand humans – their desires, intents, and states of mind – one must be able to observe and perceive certain behavioral cues. Eye gaze direction is one such cue: it is a strong form of non-verbal communication, signaling engagement, interest, and attention during social interactions. In this project, we aim to develop a large-scale gaze-tracking framework for robust 3D gaze estimation in unconstrained images, containing human subjects in indoor and outdoor environments with labeled 3D gaze across a wide range of head poses and distances.
Applications: eye tracking in public digital displays, etc.

Alternative Text



Points on Screen
Point View
Count Viewer
Log View