Mimir: Hierarchical Goal-Driven Diffusion with Uncertainty Propagation for End-to-End Autonomous Driving

2025-12-07

AI Summary

In end-to-end autonomous driving, inaccurate high-level guidance signals and excessive computational overhead of guidance modules hinder system performance. This paper proposes a goal-point–driven hierarchical diffusion-based trajectory planning framework. Its core contributions are threefold: (1) the first use of a Laplacian distribution to explicitly model goal-point uncertainty, enabling interpretable uncertainty quantification; (2) a dual-system, multi-rate guidance mechanism that jointly optimizes inference efficiency and robustness by coordinating high-level semantic guidance with low-level trajectory generation; and (3) an uncertainty propagation pathway that bridges high-level decision-making and low-level control. Evaluated on the NavHard and NavTest benchmarks, our method achieves a 20% improvement in EPDMS driving score, accelerates high-level module inference by 1.6×, and maintains original accuracy—demonstrating significant gains in both performance and efficiency without compromising precision.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Diffusion Models for Vision

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

End-to-end autonomous driving has emerged as a pivotal direction in the field of autonomous systems. Recent works have demonstrated impressive performance by incorporating high-level guidance signals to steer low-level trajectory planners. However, their potential is often constrained by inaccurate high-level guidance and the computational overhead of complex guidance modules. To address these limitations, we propose Mimir, a novel hierarchical dual-system framework capable of generating robust trajectories relying on goal points with uncertainty estimation: (1) Unlike previous approaches that deterministically model, we estimate goal point uncertainty with a Laplace distribution to enhance robustness; (2) To overcome the slow inference speed of the guidance system, we introduce a multi-rate guidance mechanism that predicts extended goal points in advance. Validated on challenging Navhard and Navtest benchmarks, Mimir surpasses previous state-of-the-art methods with a 20% improvement in the driving score EPDMS, while achieving 1.6 times improvement in high-level module inference speed without compromising accuracy. The code and models will be released soon to promote reproducibility and further development. The code is available at https://github.com/ZebinX/Mimir-Uncertainty-Driving

Problem

• Addresses inaccurate high-level guidance in autonomous driving systems
• Reduces computational overhead of complex guidance modules
• Enhances robustness with uncertainty estimation in trajectory planning

Innovation

• Estimates goal uncertainty using Laplace distribution for robustness
• Introduces multi-rate guidance to predict extended goals early
• Improves inference speed without compromising trajectory accuracy