MeanFuser: Fast One-Step Multi-Modal Trajectory Generation and Adaptive Reconstruction via MeanFlow for End-to-End Autonomous Driving

2026-02-23

AI Summary

This work proposes MeanFuser, an efficient and robust end-to-end trajectory planning framework for autonomous driving that overcomes the trade-off between coverage and performance inherent in existing discrete anchor-based methods. By modeling trajectories in a continuous space, MeanFuser employs Gaussian mixture noise to construct a continuous trajectory representation, leverages a mean flow identity (MeanFlow Identity) to circumvent numerical errors from ordinary differential equation (ODE) solvers, and integrates a lightweight adaptive reconstruction module (ARM) with dynamic attention to refine trajectory selection. Evaluated on the NAVSIM closed-loop benchmark, MeanFuser achieves state-of-the-art performance without relying on PDM Score supervision while significantly improving inference efficiency.

Technology Category

• Planning, Routing, and Scheduling: Mixed Discrete/Continuous Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Generative models have shown great potential in trajectory planning. Recent studies demonstrate that anchor-guided generative models are effective in modeling the uncertainty of driving behaviors and improving overall performance. However, these methods rely on discrete anchor vocabularies that must sufficiently cover the trajectory distribution during testing to ensure robustness, inducing an inherent trade-off between vocabulary size and model performance. To overcome this limitation, we propose MeanFuser, an end-to-end autonomous driving method that enhances both efficiency and robustness through three key designs. (1) We introduce Gaussian Mixture Noise (GMN) to guide generative sampling, enabling a continuous representation of the trajectory space and eliminating the dependency on discrete anchor vocabularies. (2) We adapt ``MeanFlow Identity" to end-to-end planning, which models the mean velocity field between GMN and trajectory distribution instead of the instantaneous velocity field used in vanilla flow matching methods, effectively eliminating numerical errors from ODE solvers and significantly accelerating inference. (3) We design a lightweight Adaptive Reconstruction Module (ARM) that enables the model to implicitly select from all sampled proposals or reconstruct a new trajectory when none is satisfactory via attention weights. Experiments on the NAVSIM closed-loop benchmark demonstrate that MeanFuser achieves outstanding performance without the supervision of the PDM Score. and exceptional inference efficiency, offering a robust and efficient solution for end-to-end autonomous driving. Our code and model are available at https://github.com/wjl2244/MeanFuser.

Problem

• trajectory planning
• anchor-guided generative models
• discrete anchor vocabularies
• trajectory distribution
• autonomous driving

Innovation

• MeanFlow
• Gaussian Mixture Noise
• Adaptive Reconstruction
• End-to-End Autonomous Driving
• Flow Matching