Abstract

This paper presents an online learning algorithm for appea- rance-based gaze estimation that allows free head movement in a ca- sual desktop environment. Our method avoids the lengthy calibration stage using an incremental learning approach. Our system keeps running as a background process on the desktop PC and continuously updates the estimation parameters by taking user’s operations on the PC mon- itor as input. To handle free head movement of a user, we propose a pose-based clustering approach that efficiently extends an appearance manifold model to handle the large variations of the head pose. The effectiveness of the proposed method is validated by quantitative perfor- mance evaluation with three users.