What-If Motion Prediction for Autonomous Driving

Abstract

Forecasting the long-term future motion of road actors is a core challenge to the deployment of safe autonomous vehicles (AVs). Viable solutions must account for both the static geometric context, such as road lanes, and dynamic social interactions arising from multiple actors. While recent deep architectures have achieved state-of-the-art performance on distance-based forecasting metrics, these approaches produce forecasts that are predicted without regard to the AV's intended motion plan. In contrast, we propose a recurrent graph-based attentional approach with interpretable geometric (actor-lane) and social (actor-actor) relationships that supports the injection of counterfactual geometric goals and social contexts. Our model can produce diverse predictions conditioned on hypothetical or "what-if" road lanes and multi-actor interactions. We show that such an approach could be used in the planning loop to reason about unobserved causes or unlikely futures that are directly relevant to the AV's intended route.

16 pages, 6 tables, 6 figures

References in corpus (14)

• Sequence to Sequence Learning with Neural Networks
• How would surround vehicles move? A Unified Framework for Maneuver Classification and Motion Prediction
• Fast and Furious: Real Time End-to-End 3D Detection, Tracking and Motion Forecasting with a Single Convolutional Net
• IntentNet: Learning to Predict Intention from Raw Sensor Data
• Scalability in Perception for Autonomous Driving: Waymo Open Dataset
• Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction
• Argoverse: 3D Tracking and Forecasting with Rich Maps
• MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Prediction
• Multiple Futures Prediction
• Generating Long-term Trajectories Using Deep Hierarchical Networks
• VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation
• Uncertainty-aware Short-term Motion Prediction of Traffic Actors for Autonomous Driving
• Learning Lane Graph Representations for Motion Forecasting
• Deep Kinematic Models for Kinematically Feasible Vehicle Trajectory Predictions

Cited by in corpus (11)

• THOMAS: Trajectory Heatmap Output with learned Multi-Agent Sampling
• Multimodal Motion Prediction with Stacked Transformers
• DenseTNT: End-to-end Trajectory Prediction from Dense Goal Sets
• Multimodal Trajectory Prediction Conditioned on Lane-Graph Traversals
• Towards trustworthy multi-modal motion prediction: Holistic evaluation and interpretability of outputs
• Decoder Fusion RNN: Context and Interaction Aware Decoders for Trajectory Prediction
• LaPred: Lane-Aware Prediction of Multi-Modal Future Trajectories of Dynamic Agents
• Scene Transformer: A unified architecture for predicting multiple agent trajectories
• A Scenario-Based Platform for Testing Autonomous Vehicle Behavior Prediction Models in Simulation
• Jointly Learning Agent and Lane Information for Multimodal Trajectory Prediction
• Risk-Aware Lane Selection on Highway with Dynamic Obstacles