A Unified Candidate Set with Scene-Adaptive Refinement via Diffusion for End-to-End Autonomous Driving

TL;DR

This paper introduces CdDrive, a scene-adaptive candidate set generation method for end-to-end autonomous driving that combines fixed vocabulary with diffusion-based refinement, improving performance in diverse scenarios.

Contribution

It proposes a novel candidate generation framework using diffusion denoising conditioned on vocabulary, enhancing trajectory diversity and accuracy in complex driving situations.

Findings

Achieves state-of-the-art results on NAVSIM datasets.

Effectively balances routine and interactive driving scenarios.

Demonstrates the benefits of scene-adaptive refinement and horizon-aware smoothing.

Abstract

End-to-end autonomous driving is increasingly adopting a multimodal planning paradigm that generates multiple trajectory candidates and selects the final plan, making candidate-set design critical. A fixed trajectory vocabulary provides stable coverage in routine driving but often misses optimal solutions in complex interactions, while scene-adaptive refinement can cause over-correction in simple scenarios by unnecessarily perturbing already strong vocabulary trajectories.We propose CdDrive, which preserves the original vocabulary candidates and augments them with scene-adaptive candidates generated by vocabulary-conditioned diffusion denoising. Both candidate types are jointly scored by a shared selection module, enabling reliable performance across routine and highly interactive scenarios. We further introduce HATNA (Horizon-Aware Trajectory Noise Adapter) to improve the smoothness and geometric continuity of diffusion candidates via temporal smoothing and horizon-aware noise modulation. Experiments on NAVSIM v1 and NAVSIM v2 demonstrate leading performance, and ablations verify the contribution of each component. Code: https://github.com/WWW-TJ/CdDrive.