Object Permanence Filter for Robust Tracking with Interactive Robots

Object permanence, which refers to the concept that objects continue to existeven when they are no longer perceivable through the senses, is a crucialaspect of human cognitive development. In this work, we seek to incorporatethis understanding into interactive robots by proposing a set of assumptionsand rules to represent object permanence in multi-object, multi-agentinteractive scenarios. We integrate these rules into the particle filter,resulting in the Object Permanence Filter (OPF). For multi-object scenarios, wepropose an ensemble of K interconnected OPFs, where each filter predictsplausible object tracks that are resilient to missing, noisy, and kinematicallyor dynamically infeasible measurements, thus bringing perceptional robustness.Through several interactive scenarios, we demonstrate that the proposed OPFapproach provides robust tracking in human-robot interactive tasks agnostic tomeasurement type, even in the presence of prolonged and complete occlusion.Webpage: https://opfilter.github.io/.