Trajectory Prediction using Equivariant Continuous Convolution

2020/10/21 by Robin Walters, Walters, Robin, Jinxi Li +3
Computer Science · Engineering · #Anomaly Detection Techniques and Applications #Artificial Intelligence (cs.AI) #Autonomous Vehicle Technology and Safety #FOS: Computer and information sciences #Machine Learning (cs.LG) #Robotics (cs.RO) #Video Surveillance and Tracking Methods #cs.AI #cs.LG #cs.RO

paper · pdf · doi:10.48550/arxiv.2010.11344

Accepted to ICLR 2021, 17 pages

openalex publication_date 2020/10/21 · arxiv created 2021/03/17 · arxiv updated 2021/03/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Trajectory prediction is a critical part of many AI applications, for example, the safe operation of autonomous vehicles. However, current methods are prone to making inconsistent and physically unrealistic predictions. We leverage insights from fluid dynamics to overcome this limitation by considering internal symmetry in real-world trajectories. We propose a novel model, Equivariant Continous COnvolution (ECCO) for improved trajectory prediction. ECCO uses rotationally-equivariant continuous convolutions to embed the symmetries of the system. On both vehicle and pedestrian trajectory datasets, ECCO attains competitive accuracy with significantly fewer parameters. It is also more sample efficient, generalizing automatically from few data points in any orientation. Lastly, ECCO improves generalization with equivariance, resulting in more physically consistent predictions. Our method provides a fresh perspective towards increasing trust and transparency in deep learning models.

Citations

• Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
• Long Short-Term Memory
• Exploiting Cyclic Symmetry in Convolutional Neural Networks
• Applying Deep Learning to Basketball Trajectories
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• Rotation Equivariance and Invariance in Convolutional Neural Networks
• Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories
• General E(2)-Equivariant Steerable CNNs
• Argoverse: 3D Tracking and Forecasting with Rich Maps
• Incorporating Symmetry into Deep Dynamics Models for Improved Generalization
• Learning to Simulate Complex Physics with Graph Networks
• Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data
• Uncertainty-aware Short-term Motion Prediction of Traffic Actors for Autonomous Driving
• SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
• Learning Lane Graph Representations for Motion Forecasting

Cited by

• 3D Equivariant Visuomotor Policy Learning via Spherical Projection

Related

• EquiDiff: A Conditional Equivariant Diffusion Model For Trajectory Prediction
• Probabilistic Symmetry for Multi-Agent Dynamics
• Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving
• Equivariant Map and Agent Geometry for Autonomous Driving Motion Prediction
• A Novel Deep Neural Network for Trajectory Prediction in Automated Vehicles Using Velocity Vector Field