Abstract
We describe a Markov chain Monte Carlo based particle fil- ter that efiectively deals with interacting targets, i.e., targets that are infiuenced by the proximity and/or behavior of other targets. Such in- teractions cause problems for traditional approaches to the data associ- ation problem. In response, we developed a joint tracker that includes a more sophisticated motion model to maintain the identity of targets throughout an interaction, drastically reducing tracker failures. The pa- per presents two main contributions: (1) we show how a Markov random field (MRF) motion prior, built on the fiy at each time step, can sub- stantially improve tracking when targets interact, and (2) we show how this can be done eficiently using Markov chain Monte Carlo (MCMC) sampling. We prove that incorporating an MRF to model interactions is equivalent to adding an additional interaction factor to the importance weights in a joint particle filter. Since a joint particle filter sufiers from exponential complexity in the number of tracked targets, we replace the traditional importance sampling step in the particle filter with an MCMC sampling step. The resulting filter deals effciently and effectively with complicated interactions when targets approach each other. We present both qualitative and quantitative results to substantiate the claims made in the paper, including a large scale experiment on a video-sequence of over 10,000 frames in length.