DiffE2E: Rethinking End-to-End Driving with a Hybrid Action Diffusion and Supervised Policy

TL;DR

DiffE2E introduces a hybrid diffusion and supervised policy framework for end-to-end autonomous driving, improving robustness and generalization in complex scenarios through structured latent spaces and multi-modal feature fusion.

Contribution

The paper presents a novel diffusion-supervision hybrid model with hierarchical feature alignment and a collaborative training paradigm for end-to-end driving.

Findings

Achieves state-of-the-art results in CARLA and NAVSIM benchmarks.

Models structured latent spaces for better trajectory distribution capture.

Enhances controllability and robustness of autonomous driving policies.

Abstract

End-to-end learning has emerged as a transformative paradigm in autonomous driving. However, the inherently multimodal nature of driving behaviors and the generalization challenges in long-tail scenarios remain critical obstacles to robust deployment. We propose DiffE2E, a diffusion-based end-to-end autonomous driving framework. This framework first performs multi-scale alignment of multi-sensor perception features through a hierarchical bidirectional cross-attention mechanism. It then introduces a novel class of hybrid diffusion-supervision decoders based on the Transformer architecture, and adopts a collaborative training paradigm that seamlessly integrates the strengths of both diffusion and supervised policy. DiffE2E models structured latent spaces, where diffusion captures the distribution of future trajectories and supervision enhances controllability and robustness. A global condition integration module enables deep fusion of perception features with high-level targets, significantly improving the quality of trajectory generation. Subsequently, a cross-attention mechanism facilitates efficient interaction between integrated features and hybrid latent variables, promoting the joint optimization of diffusion and supervision objectives for structured output generation, ultimately leading to more robust control. Experiments demonstrate that DiffE2E achieves state-of-the-art performance in both CARLA closed-loop evaluations and NAVSIM benchmarks. The proposed integrated diffusion-supervision policy offers a generalizable paradigm for hybrid action representation, with strong potential for extension to broader domains including embodied intelligence. More details and visualizations are available at \href{https://infinidrive.github.io/DiffE2E/}{project website}.