Research map / 3D vision / Depth and stereo
Geometric Estimation via Robust Subspace Recovery
Aoxiang Fan, Xingyu Jiang, Yang Wang, Junjun Jiang, Jiayi Ma
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 low-dimensional affine embedding (3-dim subspace, relative dim 3/5) instead of hyperplane fitting (6/7) (ablation) | SRE · All datasets average | proportion of failures (5px): 0.349 -> 0.079 | Improved |
| use low-dimensional affine embedding instead of hyperplane fitting (ablation) | SRE · RS remote sensing | proportion of failures (5px): 0.805 -> 0.085 | Improved |
| use low-dimensional affine embedding instead of hyperplane fitting (ablation) | SRE · All datasets average | mean geometric error (5px): 0.90 -> 0.94 | No clear effect |
| use low-dimensional affine embedding instead of hyperplane fitting (ablation) | SRE · All datasets average | runtime (ms): 69.3 -> 62.5 | Improved |
| use affine linear embedding (SRE-A) instead of fundamental-matrix DLT embedding (SRE-F) (ablation) | SRE · All datasets average | proportion of failures (5px): 0.245 -> 0.079 | Improved |
| use SRE-A instead of USAC2 | geometric estimator · All datasets average | proportion of failures (5px): 0.215 -> 0.079 | Improved |
| use SRE-A instead of RANSAC | geometric estimator · All datasets average | mean geometric error (5px): 1.48 -> 0.94 | Improved |
| use SRE-A instead of RANSAC | geometric estimator · All datasets average | runtime (ms): 330.2 -> 62.5 | Improved |
| use SRE-A instead of USAC1 | geometric estimator · All datasets average | runtime (ms): 23.4 -> 62.5 | Hurt |
| use SRE-A instead of MAGSAC | geometric estimator · RS remote sensing | proportion of failures (5px): 0.763 -> 0.085 | Improved |
| use SRE-A instead of USAC2 | geometric estimator · RS remote sensing | proportion of failures (5px): 0.227 -> 0.085 | Improved |
| use SRE-A instead of USAC2 with outlier rate raised up to 95% | geometric estimator · AdelaideRMF synthetic outliers | proportion of failures | Improved |
Closest papers by meaning
- A Subspace-Constrained Tyler's Estimator and its Applications to Structure from Motion (CVPR 2024 · similarity 0.88)
- Robust Homography Estimation via Dual Principal Component Pursuit (CVPR 2020 · similarity 0.87)
- An Efficient Solution to Non-Minimal Case Essential Matrix Estimation (TPAMI 2020 · similarity 0.87)
- Non-Rigid Point Set Registration with Robust Transformation Estimation under Manifold Regularization (AAAI 2017 · similarity 0.86)
- Revisiting Sampson Approximations for Geometric Estimation Problems (CVPR 2024 · similarity 0.86)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (CVPR 2025 · similarity 0.85)
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.
