Uncertainty-Aware Driver Trajectory Prediction at Urban Intersections
This addresses uncertainty-aware motion prediction for autonomous driving systems, representing an incremental improvement through hybrid modeling.
The paper tackles driver trajectory prediction at urban intersections by proposing a variational neural network approach that generates trajectory distributions with confidence estimates, improving prediction error of a physics-based model by 25% and identifying uncertain cases with 82% accuracy.
Predicting the motion of a driver's vehicle is crucial for advanced driving systems, enabling detection of potential risks towards shared control between the driver and automation systems. In this paper, we propose a variational neural network approach that predicts future driver trajectory distributions for the vehicle based on multiple sensors. Our predictor generates both a conditional variational distribution of future trajectories, as well as a confidence estimate for different time horizons. Our approach allows us to handle inherently uncertain situations, and reason about information gain from each input, as well as combine our model with additional predictors, creating a mixture of experts. We show how to augment the variational predictor with a physics-based predictor, and based on their confidence estimations, improve overall system performance. The resulting combined model is aware of the uncertainty associated with its predictions, which can help the vehicle autonomy to make decisions with more confidence. The model is validated on real-world urban driving data collected in multiple locations. This validation demonstrates that our approach improves the prediction error of a physics-based model by 25% while successfully identifying the uncertain cases with 82% accuracy.