LLaViDA: 一种用于显式推理和增强轨迹规划的大型语言视觉驾驶助手
LLaViDA: A Large Language Vision Driving Assistant for Explicit Reasoning and Enhanced Trajectory Planning
AI总结：
LLaViDA通过视觉-语言模型实现显式推理和轨迹规划，提升自动驾驶的轨迹规划性能。
摘要：
轨迹规划是自动驾驶中的一个基础但具有挑战性的组成部分。端到端规划器在恶劣天气、不可预测的人类行为或复杂的道路布局下经常失效，主要是因为它们缺乏强大的泛化能力或少量样本学习能力。我们提出了LLaViDA，一种大型语言视觉驾驶助手，利用视觉-语言模型（VLM）进行物体运动预测、语义定位和推理链推理，以实现自动驾驶中的轨迹规划。一个两阶段训练流程——监督微调后接轨迹偏好优化（TPO）——通过注入回归监督增强了场景理解和轨迹规划，产生了一个强大的“VLM自动驾驶轨迹规划器”。在NuScenes基准测试中，LLaViDA在开环轨迹规划任务中超越了最先进的端到端和其他最近的VLM/LLM基线，实现了平均L2轨迹误差为0.31米和碰撞率为0.10%的NuScenes测试集。本文的代码可在GitHub上获得。
英文摘要：
Trajectory planning is a fundamental yet challenging component of autonomous driving. End-to-end planners frequently falter under adverse weather, unpredictable human behavior, or complex road layouts, primarily because they lack strong generalization or few-shot capabilities beyond their training data. We propose LLaViDA, a Large Language Vision Driving Assistant that leverages a Vision-Language Model (VLM) for object motion prediction, semantic grounding, and chain-of-thought reasoning for trajectory planning in autonomous driving. A two-stage training pipeline--supervised fine-tuning followed by Trajectory Preference Optimization (TPO)--enhances scene understanding and trajectory planning by injecting regression-based supervision, produces a powerful "VLM Trajectory Planner for Autonomous Driving." On the NuScenes benchmark, LLaViDA surpasses state-of-the-art end-to-end and other recent VLM/LLM-based baselines in open-loop trajectory planning task, achieving an average L2 trajectory error of 0.31 m and a collision rate of 0.10% on the NuScenes test set. The code for this paper is available at GitHub.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
