Orbia: Do Generated Videos Form a Consistent 3D World?

Title: Orbia: Do Generated Videos Form a Consistent 3D World?

标题：Orbia：生成的视频是否构成了连贯的 3D 世界？

Computer Science > Computer Vision and Pattern Recognition arXiv:2610.20693 (cs) [Submitted on 14 Oct 2026] Title: Orbia: Do Generated Videos Form a Consistent 3D World? Authors: Dongyue Lu, Guian Fang, Hur Lim, Yonggang Wu, Yanrong Wang, Hongshan Chen, Shengju Qian, Yue Liao, Wei Chow, Lingdong Kong, Tianxin Huang, Yu Yang, Hongji Yang, Junchao Huang, Ao Liang, Yihao Wang, Zihan Wang, Rong Li, Hanlin Chen, Xin Wang, Wei Tsang Ooi, Mike Zheng Shou, Shuicheng Yan.

计算机科学 > 计算机视觉与模式识别 arXiv:2610.20693 (cs) [提交于 2026 年 10 月 14 日] 标题：Orbia：生成的视频是否构成了连贯的 3D 世界？作者：Dongyue Lu, Guian Fang, Hur Lim, Yonggang Wu, Yanrong Wang, Hongshan Chen, Shengju Qian, Yue Liao, Wei Chow, Lingdong Kong, Tianxin Huang, Yu Yang, Hongji Yang, Junchao Huang, Ao Liang, Yihao Wang, Zihan Wang, Rong Li, Hanlin Chen, Xin Wang, Wei Tsang Ooi, Mike Zheng Shou, Shuicheng Yan。

Abstract: Video generators serve as interactive worlds for exploration and simulation, yet realistic and visually appealing frames do not necessarily depict a consistent 3D world. Existing benchmarks assess selected geometric properties or use vision-language models to judge memory and consistency, leaving systematic 3D evaluation underexplored.

摘要：视频生成器可作为探索和模拟的交互式世界，但逼真且视觉上吸引人的帧并不一定能描绘出一个连贯的 3D 世界。现有的基准测试评估了选定的几何属性，或使用视觉语言模型来判断记忆和一致性，这使得系统性的 3D 评估仍未得到充分探索。

We introduce Orbia, the first benchmark dedicated to a systematic and comprehensive evaluation of 3D consistency in interactive world models. Through controlled exploration and revisit, we organize the evaluation into three dimensions: input preservation, generated-content persistence, and 3D self-consistency.

我们推出了 Orbia，这是首个致力于对交互式世界模型中的 3D 一致性进行系统且全面评估的基准。通过受控的探索和重访，我们将评估组织为三个维度：输入保持、生成内容持久性以及 3D 自我一致性。

For input preservation, registered RGB-D references and annotated anchors support appearance, geometry, and spatial stability measurements. For generated-content persistence, we compare revisits to regions absent from the input image with their first generated observations to test whether models preserve what they create.

在输入保持方面，配准的 RGB-D 参考和标注锚点支持外观、几何和空间稳定性测量。在生成内容持久性方面，我们将对输入图像中不存在区域的重访与其首次生成的观测结果进行比较，以测试模型是否能保持其所创造的内容。

For 3D self-consistency, cross-view depth and feature correspondence, regional scale agreement, and held-out Gaussian rendering test whether generated views describe the same 3D scene. Minute-long exploration and revisits further test long-term consistency.

在 3D 自我一致性方面，跨视图深度和特征对应、区域尺度一致性以及留出的高斯渲染测试了生成的视图是否描述了同一个 3D 场景。长达一分钟的探索和重访进一步测试了长期一致性。

Our construction and annotation pipeline brings together public datasets, self-captured scenes, and simulation, yielding 720 instances across indoor and outdoor environments and six exploration-and-revisit protocols. Evaluation spans 27 text-, action-, and camera-conditioned models through native-interface adapters.

我们的构建和标注流程汇集了公共数据集、自采集场景和模拟，产生了涵盖室内和室外环境的 720 个实例以及六种探索与重访协议。评估通过原生接口适配器涵盖了 27 种文本、动作和摄像机条件模型。

Our Orbia benchmark provides a systematic basis for diagnosing 3D inconsistency and developing more persistent video world models.

我们的 Orbia 基准为诊断 3D 不一致性以及开发更具持久性的视频世界模型提供了系统性基础。