Research map / 3D vision / Human pose estimation
CPPF: Towards Robust Category-Level 9D Pose Estimation in the Wild
Yang You, Ruoxi Shi, Weiming Wang, Cewu Lu
Check 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 CPPF with synthetic-object training instead of the Gao et al. sim-to-real baseline | NOCS REAL275 · NOCS REAL275 test | mAP (10° 5 cm): 17.1 -> 44.9 | Improved |
| Use CPPF instead of Gao et al. | SUN RGB-D chair evaluation · SUN RGB-D chairs, validation | mAP (60° 30 cm): 7.0 -> 17.6 | Improved |
| Use a bounding-box mask instead of an instance mask | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 35.3 | Hurt |
| Use 2° orientation bins instead of 1.5° (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 46.6 | Improved |
| Use 1° orientation bins instead of 1.5° (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 38.2 | Hurt |
| Use 4° orientation bins instead of 1.5° (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 42.7 | Hurt |
| Use 10,000 point pairs instead of 100,000 (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 27.2 | Hurt |
| Use 60,000 point pairs instead of 100,000 (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 42.8 | Hurt |
| Use 200,000 point pairs instead of 100,000 (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 44.9 -> 45.5 | Improved |
| Add auxiliary orientation classification (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 31.7 -> 44.9 | Improved |
| Add coarse-to-fine voting to filter noisy points (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 27.4 -> 44.9 | Improved |
| Use classification instead of direct regression for voting statistics (ablation) | CPPF · NOCS REAL275 test | mAP (10° 5 cm): 12.7 -> 44.9 | Improved |
Later papers that compared with this method
1 later accepted paper reports CPPF as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- CPPF++: Uncertainty-Aware Sim2Real Object Pose Estimation by Vote Aggregation (TPAMI 2024 · better on 6 of 6 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- DualPoseNet — DualPoseNet: Category-Level 6D Object Pose and Size Estimation Using Dual Pose Network With Refined Learning of Pose Consistency (ICCV 2021) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- CPPF++: Uncertainty-Aware Sim2Real Object Pose Estimation by Vote Aggregation (TPAMI 2024 · similarity 0.89)
- SchemaPose: RGB-Based Category-Level Pose Estimation with Parametric Category Schema (NeurIPS 2026 · similarity 0.87)
- GIVEPose: Gradual Intra-class Variation Elimination for RGB-based Category-Level Object Pose Estimation (CVPR 2025 · similarity 0.86)
- Rethinking Correspondence-based Category-Level Object Pose Estimation (CVPR 2025 · similarity 0.86)
- RGBD Objects in the Wild: Scaling Real-World 3D Object Learning from RGB-D Videos (CVPR 2024 · similarity 0.85)
- Universal Features Guided Zero-Shot Category-Level Object Pose Estimation (AAAI 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.
