DiffRefiner: Coarse to Fine Trajectory Planning via Diffusion Refinement with Semantic Interaction for End to End Autonomous Driving

2025-11-21

AI Summary

To address the limited flexibility and accuracy in generative trajectory prediction caused by reliance on handcrafted anchors or random noise, this paper proposes a semantic-interaction-enhanced “coarse-to-fine” two-stage trajectory planning framework. In the first stage, a Transformer-based proposal module generates semantically aware initial trajectories; in the second stage, a fine-grained diffusion denoising decoder iteratively refines these proposals by fusing multimodal sensor inputs and environmental semantic cues. The method unifies discriminative proposal generation with generative refinement, significantly improving scene consistency and driving safety of predicted trajectories. Evaluated on NAVSIM v2 and Bench2Drive benchmarks, our approach achieves 87.4 EPDMS, 87.1 DS, and 71.4 SR—outperforming all prior methods and establishing new state-of-the-art results on both benchmarks.

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
• Search and Retrieval-Augmented AI: Agentic search
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Unlike discriminative approaches in autonomous driving that predict a fixed set of candidate trajectories of the ego vehicle, generative methods, such as diffusion models, learn the underlying distribution of future motion, enabling more flexible trajectory prediction. However, since these methods typically rely on denoising human-crafted trajectory anchors or random noise, there remains significant room for improvement. In this paper, we propose DiffRefiner, a novel two-stage trajectory prediction framework. The first stage uses a transformer-based Proposal Decoder to generate coarse trajectory predictions by regressing from sensor inputs using predefined trajectory anchors. The second stage applies a Diffusion Refiner that iteratively denoises and refines these initial predictions. In this way, we enhance the performance of diffusion-based planning by incorporating a discriminative trajectory proposal module, which provides strong guidance for the generative refinement process. Furthermore, we design a fine-grained denoising decoder to enhance scene compliance, enabling more accurate trajectory prediction through enhanced alignment with the surrounding environment. Experimental results demonstrate that DiffRefiner achieves state-of-the-art performance, attaining 87.4 EPDMS on NAVSIM v2, and 87.1 DS along with 71.4 SR on Bench2Drive, thereby setting new records on both public benchmarks. The effectiveness of each component is validated via ablation studies as well.

Problem

• Enhancing diffusion-based trajectory prediction with discriminative guidance
• Improving scene compliance through fine-grained environmental alignment
• Refining coarse trajectory proposals via iterative denoising process

Innovation

• Two-stage framework combines discriminative and generative planning
• Transformer-based decoder generates coarse trajectory proposals
• Diffusion refiner iteratively denoises for fine-grained trajectory refinement