VGGT-360：几何一致的零样本全景深度估计
VGGT-360: Geometry-Consistent Zero-Shot Panoramic Depth Estimation
把 VGGT 这类多视图三维重建基础模型当作零样本深度估计的几何引擎，通过全景重投影把逐视图推理统一成一致的三维理解；其中不确定性引导的自适应投影切片与注意力结构显著性增强，可直接迁移到其他需要弥合域差距（如鱼眼、全景、广角）的视觉任务。
原文依据 reformulates the task as panoramic reprojection over multi-view reconstructed 3D models by leveraging the intrinsic 3D consistency of VGGT-like foundation models
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 VGGT-360，一种用于零样本、几何一致的全景深度估计的新型免训练框架。与以往视图独立的免训练方法不同，VGGT-360 利用 VGGT 类基础模型内在的三维一致性，将任务重新表述为对多视图重建三维模型的全景重投影，从而把碎片化的逐视图推理统一为连贯的全景理解。为实现鲁棒且准确的估计，VGGT-360 集成了三个即插即用模块，构成统一的全景到三维再到深度的框架：(i) 不确定性引导的自适应投影将全景图切片为透视视图，以弥合全景输入与 VGGT 透视先验之间的域差距。它估计基于梯度的不确定性，为几何贫乏区域分配更密集的视图，从而为 VGGT 提供几何信息丰富的输入。(ii) 结构显著性增强注意力通过向 VGGT 的注意力层注入结构感知置信度，增强其在三维重建过程中的鲁棒性，引导其关注几何可靠的区域并增强跨视图连贯性。(iii) 相关性加权三维模型校正利用注意力推断的相关性分数对重叠点重新加权，从而细化重建的三维模型，为准确的全景重投影提供一致的几何基础。大量实验表明，VGGT-360 在多种分辨率以及多样的室内和室外数据集上均优于已训练和免训练的最优方法。代码见 https://github.com/Yuanjiayii/VGGT-360。
英文原文摘要 1620 字符
This paper presents VGGT-360, a novel training-free framework for zero-shot, geometry-consistent panoramic depth estimation. Unlike prior view-independent training-free approaches, VGGT-360 reformulates the task as panoramic reprojection over multi-view reconstructed 3D models by leveraging the intrinsic 3D consistency of VGGT-like foundation models, thereby unifying fragmented per-view reasoning into a coherent panoramic understanding. To achieve robust and accurate estimation, VGGT-360 integrates three plug-and-play modules that form a unified panorama-to-3D-to-depth framework: (i) Uncertainty-guided adaptive projection slices panoramas into perspective views to bridge the domain gap between panoramic inputs and VGGT's perspective prior. It estimates gradient-based uncertainty to allocate denser views to geometry-poor regions, yielding geometry-informative inputs for VGGT. (ii) Structure-saliency enhanced attention strengthens VGGT's robustness during 3D reconstruction by injecting structure-aware confidence into its attention layers, guiding focus toward geometrically reliable regions and enhancing cross-view coherence. (iii) Correlation-weighted 3D model correction refines the reconstructed 3D model by reweighting overlapping points using attention-inferred correlation scores, providing a consistent geometric basis for accurate panoramic reprojection. Extensive experiments show that VGGT-360 outperforms both trained and training-free state-of-the-art methods across multiple resolutions and diverse indoor and outdoor datasets. The code is available at https://github.com/Yuanjiayii/VGGT-360.