GeoPano: Towards Geometrically Accurate 3D Reconstruction from a Single Panorama

Jing OU; Zidong Cao; Liaoyuan Fan; Zhuoxiao Li; Shuai Zhang; Yinrui Ren; Dongli Wu; Haoang Li

cs.CVOct 15, 2026

GeoPano: Towards Geometrically Accurate 3D Reconstruction from a Single Panorama

Authors: Jing OU, Zidong Cao, Liaoyuan Fan, Zhuoxiao Li, Shuai Zhang, Yinrui Ren, Dongli Wu, Haoang Li,

Organizations: The Hong Kong University of Science and Technology (Guangzhou), China · The Hong Kong University, Hong Kong · Sun Yat-sen University, Guangzhou, China

Abstract

Panoramic 3D reconstruction enables the holistic recovery of complete 360 $^\circ$ scene geometry, offering significant benefits for robotics and AR/VR applications. However, directly estimating geometry from ERP images is inherently challenged by spherical distortions. To reduce spherical projection distortions, a common strategy is to decompose panoramas into distortion-free perspective views, but independently processing these views often introduces cross-view geometric inconsistencies. To address these challenges, we introduce GeoPano, a feed-forward framework for single-view panoramic 3D reconstruction that recovers a geometrically accurate and consistent point cloud from a single panorama. Our approach is driven by two key innovations: First, we propose a panoramic-aware attention aggregator, featuring a streamlined attention architecture equipped with a novel Pano-Attention mechanism, which explicitly exploits the intrinsic spherical geometry of the panorama to guide cross-view interactions, thereby suppressing geometric inconsistencies. Furthermore, we formulate a hybrid supervision strategy to jointly enforce fine-grained local details in perspective space and global structural coherence in spherical space. Experiments demonstrate that our method significantly outperforms previous SOTA approaches in reconstruction accuracy, geometric consistency, and visual fidelity.

Figures & tables

Figure 1: For each scene, our GeoPano takes a single panorama as input, directly recovers distortion-free 3D point clouds, which can be easily re-projected into the sphere plane to obtain the ERP depth map. Given diverse in-the-wild panoramas from both indoor and outdoor scenes, GeoPano can robustly reconstruct high-fidelity 3D geometry, showing its impressive zero-shot capabilities.

Figure 2: Statistics of global attention cross-view interactions in $\pi^{3}$ .

Figure 3: Overview of the framework. Given a panorama, we decompose it into perspective views and encode them using DINOv2. The extracted features are then aggregated by a panoramic-aware attention aggregator . Based on the aggregated features, the model predicts local point maps, which are fused into a global point cloud. We further introduce a hybrid loss to jointly enforce local details and global consistency. G-Attn and F-Attn denote Global Attention and Frame Attention.

Figure 4: Left: View decomposition. The input panorama is projected into 10 perspective views: 8 evenly distributed horizontally, plus upward and downward views. Right: Token mask construction. Candidate masks are constructed by thresholding the pairwise cosine similarity of token ray directions.

Method		Matterport3D (Indoor)						Stanford2D3D (Indoor)						Deep360 (Outdoor)
		3D Metrics			2D Metrics			3D Metrics			2D Metrics			3D Metrics			2D Metrics
Model	Input	Chamfer $\downarrow$	F-Score $\uparrow$	IoU $\uparrow$	AbsRel $\downarrow$	RMSE $\downarrow$	$\delta_{1}$ $\uparrow$	Chamfer $\downarrow$	F-Score $\uparrow$	IoU $\uparrow$	AbsRel $\downarrow$	RMSE $\downarrow$	$\delta_{1}$ $\uparrow$	Chamfer $\downarrow$	F-Score $\uparrow$	IoU $\uparrow$	AbsRel $\downarrow$	RMSE $\downarrow$	$\delta_{1}$ $\uparrow$
PanDA	ERP	0.553	26.270	15.710	0.0888	0.3325	0.9209	0.727	15.990	8.850	0.1648	0.2364	0.7326	2.350	18.420	10.810	0.3319	10.7668	0.4498
DA 2		0.180	77.450	67.700	0.0667	0.2882	0.9561	0.145	80.310	69.360	0.0723	0.1400	0 .9545	4.036	3.520	1.920	0.1991	12.3898	0.8155
DAP		0.269	63.260	49.930	0.1186	0.7510	0.8518	0.487	33.120	20.600	0.0921	0.3820	0.9135	2.214	34.180	21.850	0.1213	7.2308	0.9159
PanoVGGT		-	-	-	-	-	-	-	-	-	-	-	-	1.537	69.094	55.587	0.1618	6.1886	0.7561

Table 1: Quantitative evaluation on 3D and 2D metrics. ERP and Pinhole denote methods using ERP projection images and pinhole images decomposed from panoramas as input, respectively.

Figure 5: Qualitative comparison in outdoor scenarios. We compare GeoPano against DA 2 , DAP, and RPG360. As illustrated, our model produces more precise 3D geometry and significantly mitigates structural distortions in long-range regions.

Figure 6: Qualitative comparison in indoor scenes. Compared to DA 2 , DAP, and RPG360, our method achieves superior global structural coherence and recovers fine-grained local geometries.

Drop Global Attention	Pano-Attention	Hybrid Loss	Chamfer $\downarrow$	F-Score $\uparrow$	IoU $\uparrow$	Inference Time (s)	GPU Mem. (GB)
$\times$	$\times$	$\times$	0.164	81.110	71.640	0.500	4.890
$\checkmark$	$\times$	$\times$	0.157	81.780	72.370	0.370	4.250
$\checkmark$	$\checkmark$	$\times$	0.155	82.010	72.710	0.450	5.170
$\checkmark$	$\checkmark$	$\checkmark$	0.153	82.250	72.920	0.450	5.170

Table 2: Ablation study of the proposed components. Each designed module effectively improves the reconstruction accuracy. Compared with the baseline, our full model reduces the inference time by 10.3%, with only a 5.6% increase in memory consumption.

Layers w/o Global Attntion	Chamfer $\downarrow$	F-Score $\uparrow$	IoU $\uparrow$	Inference Time (s)	GPU Mem. (GB)
Baseline	0.164	81.110	71.640	0.500	4.890
$0\sim 4$	0.161	81.530	71.970	0.440	4.540
$0\sim 6$	0.157	82.050	72.680	0.410	4.440
$0\sim 8$	0.157	81.780	72.370	0.390	4.250
$0\sim 10$	0.159	81.450	72.030	0.350	4.000

Table 3: Ablation study on the effect of dropping early global attention layers in alternating attention.

Figure 7: Qualitative comparison with and without Pano-Attention. Integrating our proposed Pano-Attention significantly enhances the geometric consistency.

Table 11

Panoramic 3D Reconstruction

Panoramic 3D reconstruction recovers navigable, geometrically consistent scenes from 360-degree images, video, and sparse multi-view captures.

See all 31 papers

Unified Panoramic Geometry Estimation via Multi-View Foundation Models
May 25, 2026Vukasin Bozic, Isidora Slavkovic, Dominik Narnhofer +43D Foundation Models 3D ViTs
ETH Zürich · Google · Athlence Sports +1
PanoVGGT: Feed-Forward 3D Reconstruction from Panoramic Imagery
Mar 18, 2026Yijing Guo, Mengjun Chao, Luo Wang +5Panoramic 3D Reconstruction Feed-Forward 3D Reconstruction
ShanghaiTech University · Sudo
Pano360: Perspective to Panoramic Vision with Geometric Consistency
Mar 12, 2026Zhengdong Zhu, Weiyi Xue, Zuyuan Yang +2Multi-View Consistency Multi-View Geometry
South China University of Technology · Tongji University · Guangdong University of Technology

May 25, 2026J/K move · Enter open · S save

Unified Panoramic Geometry Estimation via Multi-View Foundation Models

Vukasin Bozic, Isidora Slavkovic, Dominik Narnhofer, Nando Metzger+3

ETH Zürich · Google · Athlence Sports+1

3D Foundation Models 3D ViTs Panoramic 3D Reconstruction 3D Reconstruction Depth Estimation Single-View 3D Reconstruction+1

Open paper AskPDF arXiv

Geometry estimation from perspective images has greatly advanced, maturing to the point where off-the-shelf foundation models are able to reconstruct 3D scene structure not only from multi-view imagery, but even from a single view. A natural extension is 3D reconstruction from panoramas, with the exciting prospect of recovering a full 360-degree scene from a single panoramic image. In this work, we introduce PaGeR (Panoramic Geometry Reconstruction), a framework to lift powerful 3D foundation models designed for perspective imagery to the panorama domain. Our strategy is to start from a pre-trained transformer for 3D reconstruction and turn it into a unified high-performance model that predicts scale-invariant depth, metric depth, surface normals, and sky masks from both perspective and omnidirectional images, in a single forward pass. By keeping architectural changes to a minimum and mixing perspective and panoramic images during training, PaGeR retains the rich 3D prior of the underlying foundation model while learning to also estimate geometrically consistent 360-degree scenes from single panoramas. We extensively test our method in both indoor and outdoor environments and find that it delivers state-of-the-art performance and excellent zero-shot performance across a wide range of scenes. Code, data and models are available here.

Abstract

Panoramic 3D Reconstruction

Explore similar work

Unified Panoramic Geometry Estimation via Multi-View Foundation Models

PanoVGGT: Feed-Forward 3D Reconstruction from Panoramic Imagery

Pano360: Perspective to Panoramic Vision with Geometric Consistency