# 用于内物流中上下文语义建图的视觉-语言模型推理

> 原文链接：https://robotics.hellobike.com/papers/2026-09-01-2606-24814
> 频道：papers | 更新日期：2026-09-01 | 标签：arxiv, cs.RO
> 核心摘要：面向内物流机器人缺乏物体语义与上下文属性的问题，提出结合SLAM、SAM、聚类与VLM多视角推理的语义建图流程，并在语义分类和可移动性估计上验证。

# 用于内物流中上下文语义建图的视觉-语言模型推理

::: v-pre
> 原文链接：[Vision-Language Model Reasoning for Contextual Semantic Mapping in Intralogistics](https://arxiv.org/abs/2606.24814)（arXiv 2606.24814 · 发布日期：2026-09-01）

**Vision-Language Model Reasoning for Contextual Semantic Mapping in Intralogistics**

面向内物流机器人缺乏物体语义与上下文属性的问题，提出结合SLAM、SAM、聚类与VLM多视角推理的语义建图流程，并在语义分类和可移动性估计上验证。

## 核心要点

- 内物流移动机器人依赖几何地图定位导航，但缺少对物体及其上下文属性的语义理解。
- 方法结合SLAM几何建图、SAM实例分割、实例聚类与VLM多视角推理生成上下文语义地图。
- 通过多视角聚合与零样本开放词汇VLM查询，无需任务特定训练或预定义物体类别即可推断可移动性。
- 实验评估三种VLM和两种提示策略，语义分类达98.93% mIoU，可移动性估计达89.17% mAcc。

**验证范围与局限**：验证范围为内物流环境中的语义分类与物体可移动性估计，摘要未说明仿真或真机设置。

## 英文摘要 (Abstract)

Autonomous mobile robots operating in intralogistics environments rely on geometric maps for localization and navigation, but lack semantic understanding of objects and their contextual properties. We present a contextual semantic mapping pipeline that combines SLAM-based geometric mapping, SAM-based instance segmentation, instance clustering, and VLM multi-view reasoning to produce a contextual semantic map representation encoding geometric structure, object class, and object movability. By aggregating observations across multiple viewpoints and querying a VLM in a zero-shot, open-vocabulary setting, the pipeline infers contextual object properties--here demonstrated through movability--without requiring task-specific training or predefined object categories. We evaluate three VLMs under two prompting strategies and conduct a component-wise analysis of the pipeline. The proposed pipeline achieves 98.93 % mIoU for semantic classification and 89.17 % mAcc for object movability estimation. Component analysis identifies VLM reasoning as the primary bottleneck for contextual understanding and instance clustering as the main limitation for panoptic performance. The resulting semantic map supports context-aware filtering and robust navigation in dynamic intralogistics environments.

作者：Marvin Rüdt, Hao Pang, Constantin Enke, Zäzilia Seibold, Kai Furmans
:::

## 原始参考来源 (Sources)

- [arXiv 2606.24814：Vision-Language Model Reasoning for Contextual Semantic Mapping in Intralogistics](https://arxiv.org/abs/2606.24814)