PerceptDrive：用于端到端自动驾驶的具有自适应专家路由的感知先验世界-动作建模

中文摘要 AI 辅助

冻结的感知基础模型编码了丰富的几何、语义和动态知识。然而，狭窄的条件接口可能会削弱与任务相关的线索，而静态融合无法调整每个场景的专家贡献。我们将此挑战视为先验到计划的转移问题，并引入了PerceptDrive，这是一个具有自适应专家路由的感知先验世界-动作建模框架。PerceptDrive将来自冻结的、适应驾驶的提供者的教师提炼先验和来自冻结的自监督视频编码器的密集观察潜变量输入到一个可训练的专家路由世界-动作模型中。特定专家的查询分支处理这些信号，而先验保留目标将每个分支锚定到其先验。一个路由器从共享场景表示中预测软门，并在轨迹生成之前组合专家条件。在训练期间，基于特权规则的子度量估计为特定分支的轨迹草稿提供软门提炼目标。预测的无动作未来潜变量为流匹配演员提供条件。在推理时，没有特权组件；使用一个前置摄像头，PerceptDrive在每个规划步骤生成一个轨迹，无需测试时评分、重新排序或搜索。实验表明，PerceptDrive在NAVSIM v1上以90.4 PDMS和在NAVSIM v2上以90.2 EPDMS实现了领先的性能，优于现有方法。消融实验证实了先验保留和场景条件路由的互补收益，以及对这三种先验的不同依赖。这些结果表明，保留和自适应路由感知先验可以在无需测试时候选选择的情况下改进直接规划。

英文摘要

Frozen perception foundation models encode rich geometric, semantic, and dynamic knowledge. Yet narrow conditioning interfaces may attenuate task-relevant cues, while static fusion cannot adjust expert contributions to each scene. We cast this challenge as the prior-to-plan transfer problem and introduce PerceptDrive, a perception prior world-action modeling framework with adaptive expert routing. PerceptDrive feeds teacher-distilled priors from a frozen, driving-adapted provider and dense observation latents from a frozen self-supervised video encoder into a trainable expert-routed world-action model. Expert-specific query branches process these signals, while a prior-retention objective anchors each branch to its prior. A router predicts soft gates from a shared scene representation and combines the expert conditions before trajectory generation. During training, privileged rule-based sub-metric estimates for branch-specific trajectory drafts provide soft-gate distillation targets. The predicted action-free future latent conditions a flow-matching actor. At inference, privileged components are absent; with one front-facing camera, PerceptDrive generates one trajectory per planning step without test-time scoring, reranking, or search. Experiments show that PerceptDrive achieves state-of-the-art performance with 90.4 PDMS on NAVSIM v1 and 90.2 EPDMS on NAVSIM v2, outperforming existing methods. Ablations confirm complementary gains from prior retention and scene-conditioned routing, alongside differential reliance on the three priors. These results demonstrate that preserving and adaptively routing perception priors improves direct planning without test-time candidate selection.

发表机构

• Tsinghua University（清华大学）
• AMap, Alibaba Group（高德软件有限公司，阿里巴巴集团）

机构由 AI 辅助整理，请以论文原文为准。