Research map / 3D vision / Depth and stereo
Using Sparse Elimination for Solving Minimal Problems in Computer Vision
Janne Heikkila
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 sparse elimination basis selection instead of modified resultant basis | PEP solver for equal-focal-length self-calibration · Synthetic, sigma=0.5 px | mean relative error: 0.1947 -> 0.1935 | Improved |
| use sparse elimination basis selection instead of modified resultant basis | PEP solver for equal-focal-length self-calibration · Noise-free synthetic data | relative error distribution | Improved |
| use sparse elimination basis selection instead of modified resultant basis | PEP solver for equal-focal-length self-calibration · Real images (checkerboard, 3 images, 27 of 54 points) | relative error | Improved |
| use sparse elimination basis selection instead of modified resultant basis | PEP solver for equal-focal-length self-calibration · Real images (transparency film box, 14 images) | relative error | Improved |
| use sparse elimination instead of modified resultant basis | PEP solver · Equal focal length problem | spurious roots: 99 -> 12 | Improved |
| use sparse elimination instead of modified resultant basis | PEP solver · Equal focal length problem | execution time (ms): 17.5 -> 3.2 | Improved |
| rescale pixel values by 1/10 before computing homographies (ablation) | SPE solver · Synthetic noise sigma=0.5 px | numerical stability | Improved |
| use more points than the minimum four for homography estimation (ablation) | SPE solver · Real checkerboard images | relative error: 4 points (minimum) vs 27 points | Improved |
| remove zero monomials via SVD-based elimination (ablation) | MRE with random equation selection · Equal focal length problem | stability: removed 11 zero monomials | Improved |
| use unequal focal length problem (4 constraints instead of 3) in noisy data | SPE2 · Synthetic sigma=0.5 px | precision | Improved |
| use unequal focal length problem in noise-free data | SPE2 · Noise-free synthetic data | precision | Hurt |
| removing 15 common zero columns from C0 and C1 (ablation) | SPE2 (basis 191) · Unequal focal length problem | number of solutions: 191 -> 176 | Improved |
Closest papers by meaning
- A Sparse Resultant Based Method for Efficient Minimal Solvers (CVPR 2020 · similarity 0.93)
- Beyond Grobner Bases: Basis Selection for Minimal Solvers (CVPR 2018 · similarity 0.91)
- A Clever Elimination Strategy for Efficient Minimal Solvers (CVPR 2017 · similarity 0.89)
- Optimizing Elimination Templates by Greedy Parameter Search (CVPR 2022 · similarity 0.89)
- Automatic Solver Generator for Systems of Laurent Polynomial Equations (IJCV 2026 · similarity 0.88)
- Making Minimal Solvers for Absolute Pose Estimation Compact and Robust (ICCV 2017 · similarity 0.88)
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.
