Holo360D: A Large-Scale Real-World Dataset with Continuous Trajectories for Advancing Panoramic 3D Reconstruction and Beyond

Large-Scale Real-World Panoramic 3D Dataset

Panoramic 3D Reconstruction · Continuous Trajectories · Large-Scale Real-World Dataset

Dataset Code Paper

Demos

Click thumbnails below to view different scenes.

RGB
Valid Mask
Pointcloud Depth
Mesh Depth
Loading 3D model...
Scene 1 thumbnail
Scene 2 thumbnail
Scene 3 thumbnail
Scene 4 thumbnail
Scene 5 thumbnail

Overview

Holo360D is a large-scale real-world panoramic 3D dataset with continuous camera trajectories and accurately aligned high-completeness depth maps. The raw data are collected using a 3D laser scanner coupled with a 360-degree camera, then processed with online/offline SLAM and a panorama-specific post-processing pipeline.

109,495

Panoramas

Ground Truth

Meshes, point clouds, depth maps, and camera poses

Holo360D is the only large-scale real-world panoramic dataset that provides accurately aligned high-completeness depth maps and continuous camera trajectories.

Datasets Indoor Outdoor Continuity Alignment↓ Depth Completion↑ Scenes Numbers
Stanford2D3D ✓ × × 9.45 0.72(I) 10 1,314
Matterport3D ✓ × × 7.99 0.62(I) 90 10,790
Depth360 × ✓ ✓ (N/A) N/A N/A 30 30,000
360Loc ✓ ✓ ✓ (0.49) 12.24 0.62(I), 0.7(O) 4 2,244
KITTI-360 × ✓ ✓ (1.01) 11.72 0.16(O) 11 83,000
Holo360D (Ours) ✓ ✓ ✓ (0.29) 5.03 0.86(I), 0.82(O) 75 109,495
Paper cover figure
Holo360D provides large-scale real-world panoramic captures paired with LiDAR-derived geometry, including meshes, point clouds, depth maps, and camera poses.
Depth comparison
Compared with existing panoramic datasets, Holo360D offers more complete and better aligned depth maps for both indoor and outdoor scenes.
Algorithm comparison
Fine-tuning feed-forward 3D reconstruction models on Holo360D improves panoramic reconstruction quality and provides stronger geometric supervision.

Citation

@article{ou2026holo360d,
  title={Holo360D: A Large-Scale Real-World Dataset with Continuous Trajectories for Advancing Panoramic 3D Reconstruction and Beyond},
  author={Ou, Jing and Cao, Zidong and Ren, Yinrui and Li, Zhuoxiao and Zhu, Jinjing and Hua, Tongyan and Zhang, Shuai and Xiong, Hui and Zhao, Wufan},
  journal={arXiv preprint arXiv:2604.22482},
  year={2026}
}