No abstract

We present a novel predictive statistical framework to improve the performance of an EigenTracker. In addition, we use fast and efficient eigenspace updates to learn new views of the object being tracked on the fly. We also incorporate a new Importance Sampling mechanism which increases the robustness of the EigenTracker, and enables it to track nonconvex objects better. Our EigenTracker is flexible -it is possible to use it symbiotically with other trackers. We show its successful application in hand gesture analysis; and face and person tracking.

ARSTKACTThis paper presents a rohust shape-hascd on-line tracker for siinultanco~isly tracking the inotion of horh hands. that is robust to cases of background clutter. other inoving objects, occlusions of one hsid hy the other. and a wide range of illumination vxiatioiis. The tracker is hased on ail on-line predictive EigenTracking framework. Ibis framework allows efficient tracking of aniculate objects. which change in appearance across views. We show results o l successful tracking across all possible cases of motion dynamics of both haids during occlusion. and a widc range of illumination conditions.

Tracking highly articulated 3-D objects such as human hands with an uncalibrated camera is not an easy task. In addition to changes in position, such objects change their shape and appearance with time. We present a robust tracker for such cases which works well in spite of cases of other similar moving objects, and background clutter. Our hand gesture analysis system is based on such a tracker. The system can recognise gestures involving the same hand shapes following different trajectories, and vice versa.