User gradient-based eye center detection and head movement mapping to perform real-time gaze track.
Eye and its movement, regarding as the dominant feature of human face, is of essential importance when expressing one’s cognition, needs and emotion. Eye movement especially, is a significant way for individual to collect information from outside world. By detecting and analyzing the phenomenon of eye movement, we can apply its result to track gaze and boost the field of human–machine interaction. One more step, by comprehensive analyzing position and stop time of gaze points, we may deduce user’s pattern of browsing and contents that may interest them, then deep learning user’s personal interest and push contents that satisfies. If applied to optimize website design, these advertisement will have more pertinence and could try the new pattern of pay-per-gaze, which surly has a better publicity then pay-per-click pattern.
Keyword(s): Eye center localization, Feature extract, Gaze estimation, Shape analysis.
https://drive.google.com/open?id=0B8pACzOH-xRKTUY1QkRZUGN2aTA