Combining Behavior and Situation Information for Reliably Estimating Multiple Intentions

Intersections are the most accident-prone spots in the road network. In order to assist the driver in complex urban intersection situations, an ADAS will be required not only to recognize current but also to anticipate future maneuvers of the involved road users. Current approaches for intention estimation focus mainly on discerning only two intentions based on a vehicle's behavior. We argue that for distinguishing between more than two intentions not just a vehicle's kinematic behavior but also its driving situation needs to be taken into account. In our system we estimate four different intentions by modeling and recognizing driving situations in a Bayesian Network and using the behavior as additional evidence. For the behavior based estimation we present a newly engineered feature, the Anticipated Velocity at Stop line, that turned out to be a very strong indicator for the intention. Our system is evaluated on a real-world data set comprising approaches to seven different intersections on which we can show that our approach is able to estimate a driver's intention with a high accuracy. I. INTRODUCTION Research in the field of driving intention estimation, es- pecially associated with intersections, has become a popular topic in academia as well as for automotive manufacturers. The growing interest is due to the fact that intersections are highly accident-prone spots in the road network and a safe crossing requires to take other road users and their intentions into account. As the majority of the accidents occurring at intersections are related to driver errors, an Advanced Driver Assistance System (ADAS) that provides support and warns of critical situations would be of great benefit. In intersection scenarios especially those intentions need to be discernable that result in different maneuvers, like cross- ing straight versus turning. Knowing the path that another vehicle will take, is an important information for detecting possibly hazardous situations early on. Even estimating the intention of the ego-vehicle itself can be useful in multiple ways: For example to avoid turning accidents with parallel driving bicyclists or to prevent the driver from red-light running.