Effective Multi-Model Motion Tracking using Action Models

Yang Gu, Manuela M. Veloso

Abstract: We consider tasks where robots act on the target that is visually tracked, such as kicking a ball or pushing an object. We introduce a principled approach to incorporate models of the robot-object interaction into the tracking algorithm to effectively improve the performance of the tracker. We first present the integration of a single robot behavioral model with multiple actions into our dynamic Bayesian probabilistic tracking algorithm. We then extend to multiple motion tracking models corresponding to known multi-robot coordination plans or from multi-robot communication. We evaluate our resulting informed-tracking approach empirically in simulation and using a setup Segway robot soccer task. The input of the multiple single and multi-robot behavioral models allows a robot to visually track mobile targets with dynamic trajectories more effectively.