Situation-specific learning for ego-vehicle behavior prediction systems

We present a system able to predict the future behavior of the ego-vehicle in an inner-city environment. Our system learns the mapping between the current perceived scene (information about the ego-vehicle and the preceding vehicle, as well as information about the possible traffic lights) and the future driving behavior of the ego-vehicle. We improve the prediction accuracy by estimating the prediction confidence and by discarding unconfident samples. The behavior of the driver is represented as a sequence of elementary states termed behavior primitives. These behavior primitives are abstractions from the raw actuator states. Behavior prediction is therefore considered to be a multi-class learning problem. In this contribution, we explore the possibilities of situation-specific learning. We show that decomposing the perceived complex situation into a combination of simpler ones, each of them with a dedicated prediction, allows the system to reach a performance equivalent to a system without situation-specificity. We believe that this is advantageous for the scalability of the approach to the number of possible situations that the driver will encounter. The system is tested on a real world scenario, using streams recorded in inner-city scenes. The prediction is evaluated for a prediction horizon of 3s into the future, and the quality of the prediction is measured using established evaluation methods.

[1]  Franz Kummert,et al.  Behavior prediction at multiple time-scales in inner-city scenarios , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[2]  Anup Doshi,et al.  Lane change intent prediction for driver assistance: On-road design and evaluation , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[3]  Mohan M. Trivedi,et al.  Lane Change Intent Analysis Using Robust Operators and Sparse Bayesian Learning , 2005, IEEE Transactions on Intelligent Transportation Systems.

[4]  Rüdiger Dillmann,et al.  A probabilistic model for estimating driver behaviors and vehicle trajectories in traffic environments , 2010, 13th International IEEE Conference on Intelligent Transportation Systems.

[5]  Arati Gerdes.  Automatic Maneuver Recognition in the Automobile: the Fusion of Uncertain Sensor Values using Bayesian Models , 2006 .

[6]  Claus Möbus,et al.  Learning the human longitudinal control behavior with a modular hierarchical Bayesian Mixture-of-Behaviors model , 2011, 2011 IEEE Intelligent Vehicles Symposium (IV).

[7]  Michael Brost,et al.  Action Recognition and Prediction for Driver Assistance Systems Using Dynamic Belief Networks , 2002, Agent Technologies, Infrastructures, Tools, and Applications for E-Services.

[8]  Jannik Fritsch,et al.  Autonomous Generation of Internal Representations for Associative Learning , 2010, ICANN.

[9]  Markus Maurer,et al.  Situation aspect modelling and classification using the Scenario Based Random Forest algorithm for convoy merging situations , 2010, 13th International IEEE Conference on Intelligent Transportation Systems.

[10]  Roland Siegwart,et al.  Bayesian on-line learning of driving behaviors , 2011, 2011 IEEE International Conference on Robotics and Automation.

[11]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .