Research map / 3D vision / Human pose estimation
Relative Pose from Deep Learned Depth and a Single Affine Correspondence
Ivan Eichhardt, Daniel Barath
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 1AC+D instead of 5PT in RANSAC | GC-RANSAC · 1DSfM relative pose | processing time | Improved |
| use 1AC+D instead of 5PT in RANSAC | GC-RANSAC · 1DSfM relative pose | rotation and translation error | No clear effect |
| use 1AC+D instead of 5PT as pose-graph initializer | Theia global SfM · 1DSfM | accuracy and runtime | No clear effect |
| use 1AC+D instead of 5PT as pose-graph initializer | Theia global SfM · Vienna Cathedral | runtime (s): 81,4 -> 56,5 | Improved |
| use 1AC+D instead of 5PT as pose-graph initializer | Theia global SfM · Madrid Metropolis | orientation AVG error: 6,9 -> 5,2 | Improved |
| use 1AC+D instead of 5PT as pose-graph initializer | Theia global SfM · NYC Library | runtime (s): 5,9 -> 10,4 | Hurt |
| use 1AC+D instead of 5PT as pose-graph initializer | Theia global SfM · Ellis Island | orientation AVG error: 3,1 -> 4,4 | Hurt |
| compute rotation by the proposed algorithm instead of SVD or Gram-Schmidt | 1AC+D · relative pose | speed | Improved |
| use 1AC+D (proposed) instead of Umeyama alignment | 1AC+D · solver complexity | operations per iteration: 345 -> 133 | Improved |
Closest papers by meaning
- Relative Pose Estimation through Affine Corrections of Monocular Depth Priors (CVPR 2025 · similarity 0.93)
- Semicalibrated Relative Pose from an Affine Correspondence and Monodepth (ECCV 2024 · similarity 0.92)
- From Correspondences to Pose: Non-minimal Certifiably Optimal Relative Pose without Disambiguation (CVPR 2024 · similarity 0.89)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.88)
- FAR: Flexible Accurate and Robust 6DoF Relative Camera Pose Estimation (CVPR 2024 · similarity 0.87)
- Practical Solutions to the Relative Pose of Three Calibrated Cameras (CVPR 2025 · similarity 0.87)
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.
