# GlassFormer：使用雷达-深度融合学习实时玻璃分割

> 原文链接：https://robotics.hellobike.com/papers/2026-09-29-2609-36844
> 频道：papers | 更新日期：2026-09-29 | 标签：arxiv, cs.RO
> 核心摘要：针对透明表面易被光学感知误判为空闲空间的问题，GlassFormer融合毫米波雷达与RGB-D，在混合和低光测试划分验证实时分割。

# GlassFormer：使用雷达-深度融合学习实时玻璃分割

::: v-pre
> 原文链接：[GlassFormer: Learning Real-time Glass Segmentation using Radar-Depth Fusion](https://arxiv.org/abs/2609.36844)（arXiv 2609.36844 · 发布日期：2026-09-29）

**GlassFormer: Learning Real-time Glass Segmentation using Radar-Depth Fusion**

针对透明表面易被光学感知误判为空闲空间的问题，GlassFormer融合毫米波雷达与RGB-D，在混合和低光测试划分验证实时分割。

## 核心要点

- 透明玻璃会使RGB和深度传感器失效，导致导航系统可能将墙、门或镜面误判为空闲空间。
- 方法利用雷达在玻璃上强反射的几何线索，生成雷达引导空间先验，并通过跨模态注意力融入轻量Transformer分割网络。
- 在混合条件测试划分上达到0.88 mIoU，在低光划分上达到0.59 mIoU，较仅视觉基线更鲁棒。
- 系统在资源受限平台上保持实时性能；摘要未提及开源信息。

**验证范围与局限**：验证范围为覆盖各场景类型的混合条件测试划分和专门低光测试划分。

## 英文摘要 (Abstract)

Transparent surfaces are ubiquitous in built environments, yet they remain a persistent failure case for robotic perception. RGB cameras perceive the background behind glass rather than the surface itself, while depth sensors such as LiDAR, time-of-flight, and RGB-D often return invalid or background measurements in transparent regions. As a result, systems that rely solely on optical sensing may misinterpret glass walls, doors, or mirrors as free space, compromising safe and reliable navigation. Existing glass segmentation approaches address this by learning visual cues such as reflections, boundaries, and semantic context from RGB images. While effective under favourable lighting and viewing conditions, these cues degrade in low-light environments, under glare, or when glass surfaces are featureless or partially occluded. In this work, we propose a multimodal framework that fuses millimetre-wave radar with RGB-D sensing for real-time transparent surface segmentation. Radar reflects strongly off glass surfaces, providing a geometric cue that remains reliable precisely where vision and depth fail. We exploit this cross-modal inconsistency to generate a radar-guided spatial prior, which is integrated into a lightweight transformer-based segmentation network, GlassFormer, via cross-modal attention. We report results on a mixed-condition test split covering all scene types and a dedicated low-light split designed to stress vision-only methods. GlassFormer achieves 0.88 mIoU on the mixed split, and 0.59 mIoU on the low light split, demonstrating substantial robustness gains over vision-only baselines while maintaining real-time performance on resource-constrained platforms.

作者：Suhani Grover, Astik Srivastava, Viswas Dinesh, Avinash Sharma, K. Madhava Krishna
:::

## 原始参考来源 (Sources)

- [arXiv 2609.36844：GlassFormer: Learning Real-time Glass Segmentation using Radar-Depth Fusion](https://arxiv.org/abs/2609.36844)