Research map / 3D vision / Depth and stereo
Real-Time Sphere Sweeping Stereo From Multiview Fisheye Images
Andreas Meuleman, Hyeonjoong Jang, Daniel S. Jeon, Min H. Kim
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Select the best camera for each pixel using distance-discriminating power rather than using all views (ablation) | Sphere-sweeping stereo · Synthetic dataset, 2048×1024 px | RMSE: 0.082 -> 0.079 | Improved |
| Select the best camera for each pixel rather than use all views (ablation) | Sphere-sweeping stereo · Synthetic dataset, 2048×1024 px | Runtime (ms), 1080 Ti: 5.7·100 -> 2.7·100 | Improved |
| Use inter-scale bilateral filtering instead of the fast bilateral solver (ablation) | Sphere-sweeping stereo · Synthetic dataset, 2048×1024 px | RMSE: 0.081 -> 0.079 | Improved |
| Use inter-scale bilateral filtering instead of the fast bilateral solver (ablation) | Sphere-sweeping stereo · Synthetic dataset, 2048×1024 px | Runtime (ms), Xavier: 1.4·103 -> 1.0·101 | Improved |
| Use sphere-sweeping stereo with adaptive matching and inter-scale bilateral filtering rather than competing omnidirectional stereo methods | Ours · Synthetic dataset, 1024×512 px | RMSE: 0.168 -> 0.095 | Improved |
| Use sphere-sweeping stereo with adaptive matching and inter-scale bilateral filtering rather than competing omnidirectional stereo methods | Ours · Synthetic dataset, 1024×512 px | >0.4 inverse distance (% | m-1): 2.53 -> 0.56 | Improved |
| Use sphere-sweeping stereo with adaptive matching and inter-scale bilateral filtering rather than competing omnidirectional stereo methods | Ours · Synthetic dataset, 1024×512 px | Runtime (ms): 5.8·100 -> 2.8·100 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- OmniMVS — OmniMVS: End-to-End Learning for Omnidirectional Stereo Matching (ICCV 2019) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DirectFisheye-GS: Enabling Native Fisheye Input in Gaussian Splatting with Cross-View Joint Optimization (CVPR 2026 · similarity 0.87)
- Uniform Subdivision of Omnidirectional Camera Space for Efficient Spherical Stereo Matching (CVPR 2022 · similarity 0.86)
- Fisheye3R: Adapting Unified 3D Feed-Forward Foundation Models to Fisheye Lenses (ECCV 2026 · similarity 0.85)
- Exploring Surround-View Fisheye Camera 3D Object Detection (AAAI 2026 · similarity 0.85)
- Learned Binocular-Encoding Optics for RGBD Imaging Using Joint Stereo and Focus Cues (CVPR 2025 · similarity 0.84)
- UniK3D: Universal Camera Monocular 3D Estimation (CVPR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
