A feature-based detection and tracking system for gaze and smiling behaviours

摘要

In this study, we established a novel detection and tracking system for gaze and smiling behaviours. The system is capable of locating a human face automatically using multiple facial feature points. Feature points are then tracked using Autoregressive (AR) model based prediction and the optical flow based motion detector constrained using a facial feature model. This system has the advantage of automatically detecting the facial features and recovering the features lost during the tracking process. Eventually, a robust and efficient approach has been developed to support reliable head pose estimation and smile detection. Experimental results demonstrated the system's potentials in enhancing the learning of social communication skills.