Gray-box probabilistic occupancy mapping

Anthony Tompkins, Ransalu Senanayake, Fabio Ramos

Abstract: In order to deploy robots in previously unseen and unstructured environments, the robots should have the capacity to learn on their own and adapt to the changes in the environments. For instance, in mobile robotics, a robot should be able to learn a map of the environment from data itself without the intervention of a human to tune the parameters of the model. To this end, leveraging the latest developments in automatic machine learning (AutoML) and probabilistic programming, under the Hilbert mapping framework which can represent the occupancy of the environment as a continuous function of locations, we formulate a Bayesian framework to learn all parameters of the map. Crucially, this way, the robot is capable of learning the optimal shapes and placement of the kernels in Hilbert maps by merely embedding high-level human knowledge of the problem by means of prior probability distributions. A direct consequence of this is the ability to enable improved risk management through more robust perception and planning in complex environments. Experiments conducted on simulated and real-world datasets demonstrate the importance of incorporating prior information.

TL;DR: We present a method to automatically learn all kernel parameters for non-stationary probabilistic occupancy mapping

Keywords: occupancy mapping, autoML, probabilistic programming, nonstationary kernels