We present a probabilistic model architecture combining a layered model of human driver expertise with a cognitive map and beliefs about the driver-vehicle state to describe the effect of anticipations on driver actions.
It implements the sensory-motor system of human drivers with autonomous, goal-based attention allocation, and anticipation processes. The model has emergent properties and combines reactive with prospective behavior based on anticipated or imagined percepts obtained from a Bayesian cognitive map.
It has the ability to predict agent’s behavior, to abduct hazardous situations (what could have been the initial situation), to generate anticipatory plans, and control countermeasures preventing hazardous situations.
We demonstrated that the Bayesian-Map-extended BAD-MoB model has the ability to predict agent’s behavior, to abduct hazardous situations (what could have been the initial situation, what could be appropriate behavior), to generate anticipatory plans, and control countermeasures preventing hazardous situations. It was demonstrated that the selection of action and goal evidence has to be planned by a higher cognitive layer residing on top of the BAD-MoB model. An implementation with real expert and novice data has to follow this conceptual study.
- Anticipatory planning
- Bayesian cognitive map
- Probabilistic driver model
- Bayesian autonomous driver model
- Mixture-of-behavior model
- Visual attention allocation
- Anticipatory plans and control
- Reactive and prospective behavior
- Risk and hazardous prevention