ObjectTwin: Measuring the Asset Correspondence Gap for Capture-Anchored Digital Twins

查看原文 / View Original 编辑 / Edit

Computer Science > Computer Vision and Pattern Recognition arXiv:2610.35904 (cs) [Submitted on 30 Sep 2026] Title:ObjectTwin: Measuring the Asset Correspondence Gap for Capture-Anchored Digital Twins Authors:Zhenghe Guo View a PDF of the paper titled ObjectTwin: Measuring the Asset Correspondence Gap for Capture-Anchored Digital Twins, by Zhenghe Guo View PDF HTML (experimental)

计算机科学 > 计算机视觉与模式识别 arXiv:2610.35904 (cs) [提交于 2026 年 9 月 30 日] 标题：ObjectTwin：测量捕获锚定数字孪生的资产对应差距作者：Zhenghe Guo 查看标题为《ObjectTwin：测量捕获锚定数字孪生的资产对应差距》的论文 PDF，作者：Zhenghe Guo 查看 PDF HTML（实验性）

Abstract:Object-wise reconstruction increasingly inserts generated meshes into captured scenes as editable proxies for observed objects. This use creates a requirement absent from isolated-asset scores and pose estimates: at the captured metric state, the proxy should reproduce calibrated observations of the same instance.

摘要：基于对象的重建越来越多地将生成的网格插入到捕获的场景中，作为观察对象的编辑代理。这种用途产生了一个孤立资产评分和姿态估计所不具备的要求：在捕获的度量状态下，代理应能重现同一实例的校准观测结果。

We present ObjectTwin, a scan-grounded measurement protocol that freezes the instance, cameras, scan-projected support and recorded depth, metric anchor, and evaluation denominator, then replaces only the object representation. The resulting asset correspondence gap measures lost observation agreement relative to scan-derived visible geometry without treating the scan as complete physical ground truth.

我们提出了 ObjectTwin，这是一种基于扫描的测量协议，它固定了实例、摄像机、扫描投影支撑和记录的深度、度量锚点以及评估分母，然后仅替换对象表示。由此产生的资产对应差距衡量了相对于扫描导出的可见几何形状的观测一致性损失，且无需将扫描视为完整的物理真实值。

Across two indoor RGB-D capture domains and two generation backends, visible, anchored generated assets lose 0.302-0.497 MaskIoU under this replacement. Every tested scene-level estimate is positive, while same-input generator preference reverses across capture domains.

在两个室内 RGB-D 捕获域和两个生成后端中，可见的、锚定的生成资产在此替换下损失了 0.302-0.497 的 MaskIoU。每个测试的场景级估计均为正值，而相同输入的生成器偏好在不同捕获域之间发生逆转。

Coverage and overlap-depth analyses locate disagreement in both visible support and surface placement; the evaluated conservative pose refinement does not systematically recover it. These results establish capture correspondence as a separately measurable interface between generated-asset quality and stronger physical or task-level evaluation.

覆盖范围和重叠深度分析定位了可见支撑和表面放置方面的不一致；所评估的保守姿态细化并不能系统地恢复这种不一致。这些结果确立了捕获对应关系作为生成资产质量与更强的物理或任务级评估之间一个可单独测量的接口。