TY - CONF T1 - Learning Higher-order Transition Models in Medium-scale Camera Networks T2 - Computer Vision, 2007. ICCV 2007. IEEE 11th International Conference on Y1 - 2007 A1 - Farrell,R. A1 - David Doermann A1 - Davis, Larry S. KW - (artificial KW - approach;medium-scale KW - association KW - Bayesian KW - camera KW - cameras;video KW - framework;data KW - fusion;iterative KW - graphical KW - intelligence);optical KW - learning;incremental KW - likelihood;multicamera KW - likely KW - methods;higher KW - methods;learning KW - model KW - model;video KW - movement;probabilistic KW - network;Bayes KW - network;most KW - order KW - partition KW - problem;higher-order KW - statistics;higher-order KW - statistics;image KW - Surveillance KW - surveillance; KW - tracking;object KW - tracking;probability;video KW - transition AB - We present a Bayesian framework for learning higher- order transition models in video surveillance networks. Such higher-order models describe object movement between cameras in the network and have a greater predictive power for multi-camera tracking than camera adjacency alone. These models also provide inherent resilience to camera failure, filling in gaps left by single or even multiple non-adjacent camera failures. Our approach to estimating higher-order transition models relies on the accurate assignment of camera observations to the underlying trajectories of objects moving through the network. We addresses this data association problem by gathering the observations and evaluating alternative partitions of the observation set into individual object trajectories. Searching the complete partition space is intractable, so an incremental approach is taken, iteratively adding observations and pruning unlikely partitions. Partition likelihood is determined by the evaluation of a probabilistic graphical model. When the algorithm has considered all observations, the most likely (MAP) partition is taken as the true object trajectories. From these recovered trajectories, the higher-order statistics we seek can be derived and employed for tracking. The partitioning algorithm we present is parallel in nature and can be readily extended to distributed computation in medium-scale smart camera networks. JA - Computer Vision, 2007. ICCV 2007. IEEE 11th International Conference on M3 - 10.1109/ICCV.2007.4409203 ER -