Abstract

We derive a probabilistic framework for robust, real-time, visual tracking of previously unseen ob jects from a moving camera. The tracking problem is handled using a bag-of-pixels representation and comprises a rigid registration between frames, a segmentation and on- line appearance learning. The registration compensates for rigid motion, segmentation models any residual shape deformation and the online ap- pearance learning provides continual refinement of both the ob ject and background appearance models. The key to the success of our method is the use of pixel-wise posteriors, as opposed to likelihoods. We demon- strate the superior performance of our tracker by comparing cost function statistics against those commonly used in the visual tracking literature. Our comparison method provides a way of summarising tracking perfor- mance using lots of data from a variety of different sequences.