Towards Reconstructing Geographically Diverse Architecture with 3D Foundation Models

Computer Science > Computer Vision and Pattern Recognition arXiv:2610.22514 (cs) [Submitted on 15 Oct 2026] Title: Towards Reconstructing Geographically Diverse Architecture with 3D Foundation Models Authors: Nicky Kriplani, Yiwen Zhang, Angelina Wang, Hadar Averbuch-Elor.

计算机科学 > 计算机视觉与模式识别 arXiv:2610.22514 (cs) [提交于 2026 年 10 月 15 日] 标题：利用 3D 基础模型重建地理多样性建筑 作者：Nicky Kriplani, Yiwen Zhang, Angelina Wang, Hadar Averbuch-Elor。

Abstract: Internet-scale photo collections of real-world landmarks have driven progress in 3D computer vision yet remain highly challenging for modern 3D foundation models (3DFMs) that estimate scene structure in a single feed-forward pass.

摘要：互联网规模的现实世界地标照片集推动了 3D 计算机视觉的进步，但对于通过单次前馈传递来估计场景结构的现代 3D 基础模型（3DFMs）而言，这仍然极具挑战性。

In this work, we introduce ArchWorld, a benchmark of 3D architectural landmarks with high geographic coverage and rich metadata. We use this benchmark to conduct a thorough error analysis of current 3DFMs to understand where they fall short and how they can be improved.

在这项工作中，我们引入了 ArchWorld，这是一个具有高地理覆盖率和丰富元数据的 3D 建筑地标基准。我们利用该基准对当前的 3DFMs 进行了彻底的误差分析，以了解它们的不足之处以及如何改进。

We examine socially relevant disparities in model error and find that model performance varies by geographic region; however, unlike typical fairness regimes, the trends only partially match training data representation.

我们研究了模型误差中与社会相关的不平等现象，发现模型性能因地理区域而异；然而，与典型的公平性机制不同，这些趋势仅部分符合训练数据的代表性。

Our investigation reveals that the observed geographic disparities reflect not only differences in data representation and scene size but also differences in region-specific architectural features and background distractors that make scenes from different regions inherently vary in difficulty.

我们的调查显示，所观察到的地理差异不仅反映了数据代表性和场景规模的差异，还反映了特定区域建筑特征和背景干扰因素的差异，这些因素使得来自不同区域的场景在难度上存在本质区别。

Leveraging insights from these model failings, we introduce an intervention scheme that improves 3DFM performance while simultaneously reducing geographic disparities.

利用从这些模型缺陷中获得的见解，我们引入了一种干预方案，在提高 3DFM 性能的同时减少了地理差异。