EasyGaze: hybrid eye tracking approach for handheld mobile devices
Eye tracking technology for mobile devices has made considerable progress. However, due to limited computing capacity and the complexity of context, the traditional image feature-based technology can not extract features accurately and thus impact the performance. This paper proposes a novel approach by fusing appearance-based and feature-based eye tracking methods. Face and eye region detection were conducted to extract features, which were then used as an input to the appearance model to detect the feature points. These feature points were used to generate feature vectors, such as corner center-pupil center, by which the gaze fixation coordinates were calculated. In order to find the feature vectors with the best performance, we conducted a comparison across different vectors under different image resolution and illumination conditions, and the results showed that the average gaze fixation accuracy was achieved with 1.93 degrees of visual angle, when the image resolution was 96×48 pixels, with light sources illuminating from the front of the eye. Compared with the current methods, our method improved the accuracy of gaze fixation and was more usable.
Cite this article