Multimodal Face Detection, Head Orientation and Eye Gaze Tracking

For several applications within the human-machine-interface domain a person's face plays a key role as an information source, such as the identification, the computation of the affective state or to predict the awareness of an user. Therefore, this paper presents a multi-modal approach for finding and tracking a face and estimating the head's gaze as well as the eyes' view direction. Throughout the paper several measurements relying on two different camera inputs are introduced which can be used to form a robust computation of the head orientation and the viewing direction of a person