Probabilistic Tracking in a Metric Space

A new, exemplar-based, probabilistic paradigm for visual tracking is presented. Probabilistic mechanisms are attractive because they handle fusion of information, especially temporal fusion, in a principled manner. “Exemplars” are selected representatives of raw training data, used here to represent probabilistic mixture distributions of object configurations. Their use avoids tedious hand-construction of object models, and problems with changes of topology. Using exemplars in place of a parameterized model poses several challenges, addressed here with what we call the “Metric Mixture” (M²) approach. The M² model has several valuable properties. Principally, it provides alternatives to standard learning algorithms by allowing the use of metrics that are not embedded in a vector space. Secondly, it uses a noise model that is learned from training data. Lastly, it eliminates any need for an assumption of probabilistic pixelwise independence. Experiments demonstrate the effectiveness of the M² model in two domains: tracking walking people using chamfer distances on binary edge images, and tracking mouth movements by means of a shuffle distance.