An Integrated Non-intrusive Real Time Eye Tracker


We present a new non-intrusive real time eye tracking methodology that works under variable and realistic lighting conditions and various face orientations, with multiple people. By combining the latest techniques in appearance based object recognition and tracking with the conventional Kalman filtering based on active IR illumination, our technique is able to benefit from the strengths of different techniques and overcome their respective limitations. Experimental study shows significant improvement of our technique over the existing techniques under realistic lighting conditions, different face orientations, occlusions, and significant external illumination changes.  Details of this work may be found in our paper.  Sample eye tracking examples from our eye tracker are given below.



Eye tracking under different face poses



Eye tracking under significant variation in illumination and face poses


Eye tracking of people with glasses


Eye tracking of multiple people