Research map / 3D vision / Human pose estimation
Globally-Optimal Inlier Set Maximisation for Simultaneous Camera Pose and Feature Correspondence
Dylan Campbell, Lars Petersson, Laurent Kneip, Hongdong Li
Read the paperCheck 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 |
|---|---|---|---|
| use tighter cuboid-based upper bounds instead of weak sphere-based bounds (ablation) | GOPAC · Synthetic, 10 random 3D points, 50% 2D outliers | runtime: weakest upper bound is 50% slower than the tightest | Improved |
| add local PnP/SoftPOSIT refinement to branch-and-bound (ablation) | GOPAC · Synthetic and real data | convergence speed | Improved |
| run full globally-optimal search instead of truncating at 30s (ablation) | GOPAC truncated · Data61/2D3D | success rate (pose): 0.64 -> 0.82 | Improved |
| run full globally-optimal search instead of truncating at 30s (ablation) | GOPAC truncated · Data61/2D3D | success rate (inliers): 0.45 -> 1.00 | Improved |
| use GOPAC instead of RANSAC with 280M iterations | RANSAC · Data61/2D3D | median translation error (m): 28.5 -> 2.30 | Improved |
| use GOPAC instead of RANSAC with 280M iterations | RANSAC · Data61/2D3D | success rate (pose): 0.09 -> 0.82 | Improved |
| increase RANSAC iterations from 20 million to 280 million | RANSAC · Data61/2D3D | success rate (pose): 0.09 -> 0.09 | No clear effect |
| use GOPAC instead of SoftPOSIT and BlindPnP with torus prior | SoftPOSIT/BlindPnP · Synthetic Monte Carlo | success rate | Improved |
| highly symmetric repetitive CAD model with 75% 3D occlusion | GOPAC · Synthetic CAD structure | pose success | Hurt |
Closest papers by meaning
- Globally-Optimal Inlier Set Maximisation for Camera Pose and Correspondence Estimation (TPAMI 2018 · similarity 0.97)
- Efficient and Outlier-Robust Simultaneous Pose and Correspondence Determination by Branch-and-Bound and Transformation Decomposition (TPAMI 2021 · similarity 0.89)
- The Alignment of the Spheres: Globally-Optimal Spherical Mixture Alignment for Camera Pose Estimation (CVPR 2019 · similarity 0.88)
- Efficient Large Scale Inlier Voting for Geometric Vision Problems (ICCV 2021 · similarity 0.88)
- From Correspondences to Pose: Non-minimal Certifiably Optimal Relative Pose without Disambiguation (CVPR 2024 · similarity 0.87)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (CVPR 2025 · 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.
