LAP: Fast LAtent Diffusion Planner for Autonomous Driving

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Jinhao, Xia, Wenlong, Zhou, Zhexuan, Song, Haoming, Gong, Youmin, Mei, Jie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916071184793600
author Zhang, Jinhao
Xia, Wenlong
Zhou, Zhexuan
Song, Haoming
Gong, Youmin
Mei, Jie
author_facet Zhang, Jinhao
Xia, Wenlong
Zhou, Zhexuan
Song, Haoming
Gong, Youmin
Mei, Jie
contents Diffusion models have demonstrated strong capabilities for modeling human-like driving behaviors in autonomous driving, but their iterative sampling process induces substantial latency, and operating directly on raw trajectory points forces the model to spend capacity on low-level kinematics, rather than high-level multi-modal semantics. To address these limitations, we propose LAtent Planner (LAP), a framework that plans in a VAE-learned latent space that disentangles high-level intents from low-level kinematics, enabling our planner to capture rich, multi-modal driving strategies. To bridge the representational gap between the high-level semantic planning space and the vectorized scene context, we introduce an intermediate feature alignment mechanism that facilitates robust information fusion. Notably, LAP can produce high-quality plans in one single denoising step, substantially reducing computational overhead. Through extensive evaluations on the large-scale nuPlan benchmark, LAP achieves state-of-the-art closed-loop performance among learning-based planning methods, while demonstrating an inference speed-up of at most 10x over previous SOTA approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2512_00470
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LAP: Fast LAtent Diffusion Planner for Autonomous Driving
Zhang, Jinhao
Xia, Wenlong
Zhou, Zhexuan
Song, Haoming
Gong, Youmin
Mei, Jie
Robotics
Diffusion models have demonstrated strong capabilities for modeling human-like driving behaviors in autonomous driving, but their iterative sampling process induces substantial latency, and operating directly on raw trajectory points forces the model to spend capacity on low-level kinematics, rather than high-level multi-modal semantics. To address these limitations, we propose LAtent Planner (LAP), a framework that plans in a VAE-learned latent space that disentangles high-level intents from low-level kinematics, enabling our planner to capture rich, multi-modal driving strategies. To bridge the representational gap between the high-level semantic planning space and the vectorized scene context, we introduce an intermediate feature alignment mechanism that facilitates robust information fusion. Notably, LAP can produce high-quality plans in one single denoising step, substantially reducing computational overhead. Through extensive evaluations on the large-scale nuPlan benchmark, LAP achieves state-of-the-art closed-loop performance among learning-based planning methods, while demonstrating an inference speed-up of at most 10x over previous SOTA approaches.
title LAP: Fast LAtent Diffusion Planner for Autonomous Driving
topic Robotics
url https://arxiv.org/abs/2512.00470