A Self-Learned Arbitration Between Model-Based and Model-Free NavigationStrategies in Autonomous Driving
Abstract
Neuroscience research shows that mammals use two systems in spatial navigation: a flexible model-based strategy and aspontaneous model-free strategy. Mammals shift from model-based to model-free strategy as skills become ”habitized”and mostly use model-based strategy when high-level planning is necessary. Inspired by this line of work, the present studyproposes a model with a novel arbitration structure that solves the navigation problem in the autonomous-driving domain.This model takes into account the information from a model-based mapping/planning system and a model-free reactivecontroller, and adopts a learning-based gating method to adaptively arbitrate between the two systems. Experiments showthat the agent generally uses the reactive system when following lanes and driving through familiar intersections, andtend to rely on the planning system at unfamiliar intersections to get information about turning directions. The results aresimilar to mammal behaviors and provide insight for autonomous driving in the real world.