Research map / 3D vision / Human pose estimation
EPOS: Estimating 6D Pose of Objects With Symmetries
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Represent objects with surface fragments rather than direct single-fragment object-coordinate regression (ablation) | EPOS · T-LESS, YCB-V, LM-O | AR: 17.2 -> 47.6; 41.7 -> 69.6; 26.8 -> 44.3 | Helped |
| Increase surface fragments from 1 to 64 (ablation) | EPOS · T-LESS | AR: 17.2 -> 47.6 | Helped |
| Increase surface fragments from 1 to 64 (ablation) | EPOS · YCB-V | AR: 41.7 -> 69.6 | Helped |
| Increase surface fragments from 1 to 64 (ablation) | EPOS · LM-O | AR: 26.8 -> 44.3 | Helped |
| Increase surface fragments from 64 to 256 (ablation) | EPOS · T-LESS | AR: 47.6 -> 45.6 | Hurt |
| Increase surface fragments from 64 to 256 (ablation) | EPOS · YCB-V | AR: 69.6 -> 71.4 | Helped |
| Increase surface fragments from 64 to 256 (ablation) | EPOS · LM-O | AR: 44.3 -> 46.0 | Helped |
| Regress precise 3D fragment coordinates instead of using fragment centers (ablation) | EPOS with 64 fragments · T-LESS | AR: 29.4 -> 47.6 | Helped |
| Regress precise 3D fragment coordinates instead of using fragment centers (ablation) | EPOS with 64 fragments · YCB-V | AR: 41.5 -> 69.6 | Helped |
| Regress precise 3D fragment coordinates instead of using fragment centers (ablation) | EPOS with 64 fragments · LM-O | AR: 31.0 -> 44.3 | Helped |
| Replace GC-RANSAC with OpenCV RANSAC (ablation) | EPOS pose fitting · T-LESS, YCB-V, LM-O | AR: 35.5 vs 47.6; 67.2 vs 69.6; 41.2 vs 44.3 | Hurt |
| Use GC-RANSAC with EPnP followed by Levenberg-Marquardt rather than DLS-PnP (ablation) | EPOS pose fitting · T-LESS, YCB-V, LM-O | AR: 44.3 -> 47.6; 67.5 -> 69.6; 35.6 -> 44.3 | Helped |
Closest papers by meaning
- ES6D: A Computation Efficient and Symmetry-Aware 6D Pose Regression Framework (CVPR 2022 · similarity 0.90)
- Pos3R: 6D Pose Estimation for Unseen Objects Made Easy (CVPR 2025 · similarity 0.88)
- SingRef6D: Monocular Novel Object Pose Estimation with a Single RGB Reference (NeurIPS 2025 · similarity 0.85)
- UA-Pose: Uncertainty-Aware 6D Object Pose Estimation and Online Object Completion with Partial References (CVPR 2025 · similarity 0.85)
- One2Any: One-Reference 6D Pose Estimation for Any Object (CVPR 2025 · similarity 0.85)
- Any6D: Model-free 6D Pose Estimation of Novel Objects (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.