arxiv.org·2026/10/26
为什么重要 将 VLM 轨迹作为几何候选偏好，并融合 MUTCD 安全约束，可改善自动驾驶工作区通行路径选择与合规性。
临时交通控制引导车辆通过工作区，要求自动驾驶车辆将可见设备的解读与连续的通行路径联系起来。几何预测器可以提供多条可行路径，但选择与临时控制一致的那一条仍然困难。我们提出 SAGE（Semantic-to-Action Grounded Execution），这是一个 read-decide-execute-check 框架，在共享候选空间内将路径构建、路径偏好和控制合规性分离。
核心思想是使用视觉语言模型（VLM）的轨迹作为对完整几何候选的偏好。角色解耦的生成器构建候选，排名融合将每个候选的几何得分与其与 VLM 轨迹的一致性相结合。围绕《统一交通控制设备手册》（MUTCD）组织的监督训练 SAGE-VLM，使其在给出通行决策和轨迹之前说明车道状态和危险侧。
预测的危险侧进入由 MUTCD 条款推导的图像空间谓词，这些谓词支持在得分和位移上受约束的最终干预。在 2,306 个真实世界 ROADWork 验证片段上，候选路径在 87% 的片段中包含与人类演示相差 10 px 以内的路径，但选择遗憾占几何排名器误差的 53%。SAGE 将平均位移误差和最终位移误差相对于排名器分别降低 8.3% 和 12.0%（主要发生在远场），并将硬约束违规降低 37%。轨迹意见改善了所有四种测试的生成器-选择器组合。
SAGE 在准确率上与最先进的 VLM 模块相当，并具有响应更快的 VLM，同时在位移误差和安全指标上优于其他四种经过适配的驾驶 VLM。项目页面：https://hhjj233.github.io/SAGE-page/
Temporary traffic control redirects vehicles through work zones, requiring automated vehicles to connect an interpretation of the visible devices to a continuous passing path.
Geometric predictors can provide several plausible paths, yet selecting the one consistent with the temporary control remains difficult.
We propose SAGE (Semantic-to-Action Grounded Execution), a read-decide-execute-check framework that separates path construction, path preference, and control compliance within a shared candidate space.
The central idea is to use a vision-language model's (VLM's) trajectory as a preference over intact geometric candidates.
A role-decoupled generator constructs the candidates, and rank fusion combines each candidate's geometric score with its agreement with the VLM trajectory.
Supervision structured around the Manual on Uniform Traffic Control Devices (MUTCD) trains SAGE-VLM to state the lane status and hazard side before its passage decision and trajectory.
The predicted hazard side enters image-space predicates derived from MUTCD provisions, which support a final intervention bounded in score and displacement.
On 2,306 real-world ROADWork validation episodes, the candidates include a path within 10 px of the human demonstration in 87% of episodes, yet selection regret accounts for 53% of the geometric ranker's error.
SAGE reduces average and final displacement errors by 8.3% and 12.0% relative to the ranker, mostly in the far field, and hard-constraint violations by 37%.
The trajectory opinion improves all four tested generator-selector combinations.
SAGE is comparable to state-of-the-art VLM modules in accuracy, with a faster-responding VLM, and outperforms four other adapted driving VLMs in displacement error and safety measures.
Project page: https://hhjj233.github.io/SAGE-page/.
原版排版与图表 · 手机无法内嵌时可在浏览器打开。PDF 插件内选文不能传给网页助手。