Research map / 3D vision / Human pose estimation
SAR-Net: Shape Alignment and Recovery Network for Category-Level 6D Object Pose and Size Estimation
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 |
|---|---|---|---|
| use synthetic-only SAR-Net instead of DualPose | category-level COPSE · REAL275 | 5°2cm mAP: 29.3 -> 31.6 | Helped |
| use synthetic-only SAR-Net instead of DualPose | category-level COPSE · REAL275 | 5°5cm accuracy: 50.1 -> 54.9 | Helped |
| add symmetric correspondence to shape alignment (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 15.9 -> 17.8 | Helped |
| concatenate observed and symmetric point clouds into a coarse shape (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 17.8 -> 20.5 | Helped |
| centralize the concatenated point cloud (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 20.5 -> 24.1 | Helped |
| replace translation-offset voting with direct center regression (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 21.1 | Hurt |
| replace symmetric correspondence with direct shape completion (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 19.2 | Hurt |
| reduce template point count from 36 to 16 (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 22.8 | Hurt |
| increase template point count from 36 to 128 (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 21.5 | Hurt |
| replace shape-alignment rotation with quaternion (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 20.8 | Hurt |
| replace shape-alignment rotation with SVD (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 17.8 | Hurt |
| replace shape-alignment rotation with R6d (ablation) | SAR-Net · REAL275 | 5°2cm mAP: 24.1 -> 21.7 | Hurt |
Closest papers by meaning
- CAP-Net: A Unified Network for 6D Pose and Size Estimation of Categorical Articulated Parts from a Single RGB-D Image (CVPR 2025 · similarity 0.90)
- Unsupervised Learning of Category-Level 3D Pose from Object-Centric Videos (CVPR 2024 · similarity 0.87)
- SchemaPose: RGB-Based Category-Level Pose Estimation with Parametric Category Schema (NeurIPS 2026 · similarity 0.87)
- UA-Pose: Uncertainty-Aware 6D Object Pose Estimation and Online Object Completion with Partial References (CVPR 2025 · similarity 0.87)
- GIVEPose: Gradual Intra-class Variation Elimination for RGB-based Category-Level Object Pose Estimation (CVPR 2025 · similarity 0.87)
- GeoReF: Geometric Alignment Across Shape Variation for Category-level Object Pose Refinement (CVPR 2024 · 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.