CS 231A Final Project Real-time Webcam Gaze-tracking

Gaze-tracking is a useful technology that opens new avenues for HCI and research into human behavioral patterns. However, the image quality of current web cameras and the need for real-time image processing makes specialized hardware the norm in commercial offerings. We present a machine-learning based approach using well-engineered visual features to enable an accurate estimate of the user’s gaze position using normal webcam hardware, without the need for mounting on the user. Unlike other webcam-based approaches, we also model the user’s head orientation to allow the user greater flexibility in viewing postures. Our successful experiment, with a median error of 3.22° in determining the user’s gaze position, validates our approach.