Research map / 3D vision / Human pose estimation
Keypoint-Graph-Driven Learning Framework for Object Pose 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 |
|---|---|---|---|
| Add the adaptation layer and GCN structure constraint, training with synthetic and unlabeled real images | DAKDN on LINEMOD · LINEMOD | ADD accuracy: 46.5 -> 68.2 | Helped |
| Add unlabeled real images to synthetic-image training (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: 46.5 -> 51.8 | Helped |
| Add the GCN structure constraint (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: 46.5 -> 50.4 | Helped |
| Add GCN geometry information (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: +3.5% | Helped |
| Add GCN feature information (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: +1.6% | Helped |
| Add both GCN geometry and feature information rather than using either alone (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: 53.4 -> 68.2; 55.3 -> 68.2 | Helped |
| Replace HRNet with Stacked Hourglass in the proposed training framework (ablation) | DAKDN on LINEMOD · LINEMOD | ADD accuracy: 31.3 -> 46.1 | Helped |
| Use the keypoint-graph-driven framework instead of DPOD | 6D pose estimation · LINEMOD | ADD(-S) mean accuracy: 66.4 -> 68.2 | Helped |
| Use the keypoint-graph-driven framework instead of Self6D | 6D pose estimation · HomebrewedDB | Average Recall(%): 59.7 -> 63.8 | Helped |
| Use the keypoint-graph-driven framework instead of Self6D | 6D pose estimation · OCCLUSION | ADD(-S) Average Recall(%): 32.1 -> 33.7 | Helped |
| Remove the GCN block (ablation) | Keypoint-graph-driven framework on OCCLUSION · OCCLUSION | Average Recall(%): 33.7 -> 12.5 | Hurt |
Closest papers by meaning
- Learning Deep Network for Detecting 3D Object Keypoints and 6D Poses (CVPR 2020 · similarity 0.92)
- KeyPose: Category-Level 6D Object Pose Estimation with Self-Adaptive Keypoints (AAAI 2025 · similarity 0.88)
- Instance-Adaptive and Geometric-Aware Keypoint Learning for Category-Level 6D Object Pose Estimation (CVPR 2024 · similarity 0.87)
- FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects (CVPR 2024 · similarity 0.87)
- One2Any: One-Reference 6D Pose Estimation for Any Object (CVPR 2025 · similarity 0.86)
- Any6D: Model-free 6D Pose Estimation of Novel Objects (CVPR 2025 · similarity 0.86)
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.