Robust Facial Feature Tracking Using an Autoregressive Model and Optical Flow

A robust facial feature tracking system using the Autoregressive model based prediction and the optical flow based motion detector is proposed in this paper to track six facial features, i.e., two pupils, nostrils and lip corners. In order to improve the robustness of the tracking system, a simple facial feature model was employed to estimate the relative face poses. This system has the advantage of automatically detecting the facial features and recovering the features lost during the tracking process by re-initialization. Encouraging results have been obtained using the proposed system.