Generalizations of the auxiliary particle filter for multiple target tracking

This paper introduces two generalizations of the celebrated auxiliary particle filter for multiple target tracking. The inherent difficulty of this problem is caused by the sampling of a high dimension state space, giving rise to the curse of dimensionality, which pulls down the performance of direct generalizations of single target particle filter algorithms. The two proposed particle filters are tested in a demanding multiple target scenario, exhibiting a considerable performance improvement with respect to previously reported algorithms of this type for multiple target tracking.