CoPlanner: An Interactive Motion Planner with Contingency-Aware Diffusion for Autonomous Driving
This is the official PyTorch implementation of the paper: "CoPlanner: An Interactive Motion Planner with Contingency-Aware Diffusion for Autonomous Driving", accepted at ICRA 2026.
- [2026-02]: 🎉 CoPlanner has been accepted to ICRA 2026!
- [2026-02]: Initial code release including Model Training (Part 1 & 2) and Inference.
CoPlanner leverages a contingency-aware diffusion framework to handle interactive scenarios in autonomous driving. It generates diverse trajectory candidates by accounting for the multi-modal future behaviors of other agents.
📝 Roadmap / TODO
- Training Code (Part 1 & Part 2)
- Core Inference Engine
- Post-processing & Refinement Module
- Pre-trained Model Weights
- Visualization Tools
📑 Citation If you find our work useful in your research, please consider citing:
Code snippet @inproceedings{yourname2026coplanner, title={CoPlanner: An Interactive Motion Planner with Contingency-Aware Diffusion for Autonomous Driving}, author={Your Name and Co-authors}, booktitle={2026 IEEE International Conference on Robotics and Automation (ICRA)}, year={2026} }