ZhengYinan-AIR/Diffusion-Planner
PyTorch implementation of a diffusion-based motion planning system for autonomous vehicles.
Not currently ranked — collecting fresh signals.
star history
This repository implements a diffusion model approach for autonomous driving motion planning, published at ICLR 2025 as an oral presentation. The method uses denoising diffusion processes to generate vehicle trajectories without heavy reliance on refinement stages, enabling flexible guidance for planning in complex driving scenarios. It includes closed-loop evaluation on the nuPlan dataset and supports imitation learning integration.
Frequently asked
- What is ZhengYinan-AIR/Diffusion-Planner?
- PyTorch implementation of a diffusion-based motion planning system for autonomous vehicles.
- Is Diffusion-Planner open source?
- Yes — ZhengYinan-AIR/Diffusion-Planner is an open-source project tracked on heatdrop.
- What language is Diffusion-Planner written in?
- ZhengYinan-AIR/Diffusion-Planner is primarily written in Python.
- How popular is Diffusion-Planner?
- ZhengYinan-AIR/Diffusion-Planner has 1k stars on GitHub.
- Where can I find Diffusion-Planner?
- ZhengYinan-AIR/Diffusion-Planner is on GitHub at https://github.com/ZhengYinan-AIR/Diffusion-Planner.