Trajectory Forecasts in Unknown Environments Conditioned on Grid-Based Plans

Abstract

We address the problem of forecasting pedestrian and vehicle trajectories in unknown environments, conditioned on their past motion and scene structure. Trajectory forecasting is a challenging problem due to the large variation in scene structure and the multimodal distribution of future trajectories. Unlike prior approaches that directly learn one-to-many mappings from observed context to multiple future trajectories, we propose to condition trajectory forecasts on plans sampled from a grid based policy learned using maximum entropy inverse reinforcement learning (MaxEnt IRL). We reformulate MaxEnt IRL to allow the policy to jointly infer plausible agent goals, and paths to those goals on a coarse 2-D grid defined over the scene. We propose an attention based trajectory generator that generates continuous valued future trajectories conditioned on state sequences sampled from the MaxEnt policy. Quantitative and qualitative evaluation on the publicly available Stanford drone and NuScenes datasets shows that our model generates trajectories that are diverse, representing the multimodal predictive distribution, and precise, conforming to the underlying scene structure over long prediction horizons.

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• IntentNet: Learning to Predict Intention from Raw Sensor Data
• TNT: Target-driveN Trajectory Prediction
• Argoverse: 3D Tracking and Forecasting with Rich Maps
• Conditional Flow Variational Autoencoders for Structured Sequence Prediction
• From Goals, Waypoints & Paths To Long Term Human Trajectory Forecasting
• Map-Adaptive Goal-Based Trajectory Prediction
• TrackMPNN: A Message Passing Graph Neural Architecture for Multi-Object Tracking
• Probabilistic Multi-modal Trajectory Prediction with Lane Attention for Autonomous Vehicles

Cited by in corpus (19)

• Stepwise Goal-Driven Networks for Trajectory Prediction
• Pedestrian Models for Autonomous Driving Part II: High-Level Models of Human Behavior
• THOMAS: Trajectory Heatmap Output with learned Multi-Agent Sampling
• Energy-based Legged Robots Terrain Traversability Modeling via Deep Inverse Reinforcement Learning
• From Goals, Waypoints & Paths To Long Term Human Trajectory Forecasting
• Generating Reliable and Efficient Predictions of Human Motion: A Promising Encounter between Physics and Neural Networks
• Predicting Like A Pilot: Dataset and Method to Predict Socially-Aware Aircraft Trajectories in Non-Towered Terminal Airspace
• TrackMPNN: A Message Passing Graph Neural Architecture for Multi-Object Tracking
• MixNet: Structured Deep Neural Motion Prediction for Autonomous Racing
• Multimodal Trajectory Prediction Conditioned on Lane-Graph Traversals
• PLOP: Probabilistic poLynomial Objects trajectory Planning for autonomous driving
• BiTraP: Bi-directional Pedestrian Trajectory Prediction with Multi-modal Goal Estimation
• Trajectory Prediction with Latent Belief Energy-Based Model
• Spectral Temporal Graph Neural Network for Trajectory Prediction
• Sliding Sequential CVAE with Time Variant Socially-aware Rethinking for Trajectory Prediction
• Joint Learning Architecture for Multiple Object Tracking and Trajectory Forecasting
• From Recognition to Prediction: Analysis of Human Action and Trajectory Prediction in Video
• Evaluation metrics for behaviour modeling
• Jointly Learning Agent and Lane Information for Multimodal Trajectory Prediction