Research map / 3D vision / Depth and stereo
Structure-From-Motion Revisited
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use COLMAP's full reconstruction pipeline | Incremental SfM · Rome | registered images: 14,797 -> 20,918 | Helped |
| Use COLMAP's full reconstruction pipeline | Incremental SfM · Gendarmenmarkt | registered images: 807 -> 861 | Helped |
| Use COLMAP's full reconstruction pipeline | Incremental SfM · Quad | median camera-location error: 0.89m -> 0.85m | Helped |
| Apply recursive RANSAC triangulation instead of Bundler triangulation | Feature-track triangulation on Dubrovnik · Dubrovnik | triangulated points: 713,824 -> 902,844 | Helped |
| Use recursive RANSAC instead of recursive exhaustive pairwise sampling (ablation) | Feature-track triangulation · Dubrovnik | number of samples: 145.22 M -> 12.69 M | Helped |
| Disable recursive recovery of multiple points from merged tracks (ablation) | RANSAC feature-track triangulation · Dubrovnik | triangulated points: 861,591 -> 902,844 | Hurt |
| Choose views using a multi-resolution point-distribution score instead of point count or visible-point ratio (ablation) | Incremental reconstruction · Quad | median camera-location error | Helped |
| Group redundant views in bundle adjustment (ablation) | Bundle adjustment on a dense photo collection · Dense photo collection | pipeline runtime: 36% | Helped |
| Group redundant views in bundle adjustment at V = 0.4 (ablation) | Bundle adjustment on a dense photo collection · Dense photo collection | average reprojection error: 0.26px -> 0.28px | Hurt |
| Lower the enforced scene overlap in grouped bundle adjustment (ablation) | Bundle adjustment on a dense photo collection · Dense photo collection | pipeline runtime speedup: 5% (V = 0.6), 14% (V = 0.3) and 32% (V = 0.1) | Helped |
Closest papers by meaning
- Structure-from-Motion with a Non-Parametric Camera Model (CVPR 2025 · similarity 0.88)
- VGGSfM: Visual Geometry Grounded Deep Structure From Motion (CVPR 2024 · similarity 0.88)
- Leveraging Camera Triplets for Efficient and Accurate Structure-from-Motion (CVPR 2024 · similarity 0.88)
- Detector-Free Structure from Motion (CVPR 2024 · similarity 0.87)
- Light3R-SfM: Towards Feed-forward Structure-from-Motion (CVPR 2025 · similarity 0.87)
- Non-Rigid Structure-from-Motion: Temporally-Smooth Procrustean Alignment and Spatially-Variant Deformation Modeling (CVPR 2024 · similarity 0.86)
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.