Cross Entropy Data Association for Multiple Target Tracking

Multiple-target tracking (MTT) in the presence of spurious measurements poses difficult computational challenges related to the measurement-to-track data association problem. Different solution approaches have been proposed to tackle this problems, including various approximations and heuristic optimization tools. The Cross Entropy method is a recent optimization heuristic that has proved useful in many combinatorial optimization problems. In this work we apply the Cross-Entropy method to solve the multi-scan data association problem. We formulate the algorithm, and explore via simulation its efficiency and performance compared to other recently proposed algorithms.