Research map / 3D vision / Depth and stereo
Tangent Sampson Error: Fast Approximate Two-view Reprojection Error for Central Camera Models
Mikhail Terekhov, Viktor Larsson
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 Tangent Sampson error instead of pinhole symmetric epipolar distance in pose refinement for very wide FoV cameras | Pose refinement · FoV > 150 deg | mean rotation error (deg): 2.596 -> 1.465 | Improved |
| use Tangent Sampson error instead of pinhole Sampson in RANSAC on fisheye images | LO-RANSAC · Grossmunster | success rate (%): 74.5 -> 82.2 | Improved |
| use Tangent Sampson error instead of projective symmetric epipolar distance in RANSAC | LO-RANSAC · Grossmunster | runtime (ms): 211.2 -> 46.4 | Improved |
| use Tangent Sampson error instead of cosine error in RANSAC | LO-RANSAC · Kirchenge | median pose error (deg): 0.254 -> 0.248 | No clear effect |
| use Tangent Sampson error for low field-of-view cameras | Pose refinement · FoV < 90 deg | mean rotation error (deg): 0.376 vs 0.376 | No clear effect |
| use Tangent Sampson error versus true reprojection error optimization in refinement of > 150 deg cameras | Pose refinement · FoV > 150 deg | mean rotation error (deg): 1.418 -> 1.465 | Mixed |
| add measurement covariance to Tangent Sampson error (ablation) | Tangent Sampson · synthetic spherical cameras with unequal noise | pose error (deg) | Improved |
| use Tangent Sampson error at large initial pose error | Pose refinement · convergence test | success rate (%) | Improved |
Closest papers by meaning
- Revisiting Sampson Approximations for Geometric Estimation Problems (CVPR 2024 · similarity 0.88)
- Neural Reprojection Error: Merging Feature Learning and Camera Pose Estimation (CVPR 2021 · similarity 0.86)
- On the Instability of Relative Pose Estimation and RANSAC's Role (CVPR 2022 · similarity 0.86)
- BabelCalib: A Universal Approach to Calibrating Central Cameras (ICCV 2021 · similarity 0.85)
- LoRA3D: Low-Rank Self-Calibration of 3D Geometric Foundation models (ICLR 2025 · similarity 0.84)
- RUBIK: A Structured Benchmark for Image Matching across Geometric Challenges (CVPR 2025 · similarity 0.83)
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.
