Research map / 3D vision / Human pose estimation
Pose for Everything: Towards Category-Agnostic Pose Estimation
Lumin Xu, Sheng Jin, Wang Zeng, Wentao Liu, Chen Qian, Wanli Ouyang, Ping Luo, Xiaogang Wang
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 |
|---|---|---|---|
| add Keypoint Interaction Module (self and cross attention) (ablation) | POMNet · MP-100 Split1 1-shot | PCK: 74.40 -> 84.23 | Improved |
| add self-attention among keypoints (ablation) | POMNet · MP-100 Split1 1-shot | PCK: 80.76 -> 84.23 | Improved |
| add cross-attention between support and query (ablation) | POMNet · MP-100 Split1 1-shot | PCK: 82.92 -> 84.23 | Improved |
| add matching head instead of matrix multiplication (ablation) | POMNet · MP-100 Split1 1-shot | PCK: 79.19 -> 84.23 | Improved |
| add self-attention only on top of matching head (ablation) | POMNet · MP-100 Split1 1-shot | PCK: 74.40 -> 82.92 | Improved |
| increase number of training categories from 1 to 99 (ablation) | POMNet · human body held out, 1-shot | PCK: 39.32 -> 73.82 | Improved |
| reduce support from 5-shot to 1-shot | POMNet · MP-100 mean | PCK: 80.71 -> 79.70 | No clear effect |
| reduce support from 5-shot to 1-shot | ProtoNet · MP-100 mean | PCK: 58.56 -> 44.78 | Hurt |
| POMNet vs Fine-tune baseline | POMNet · MP-100 5-shot mean | PCK: 64.61 -> 80.71 | Improved |
| POMNet vs best baseline on vehicle super-category | POMNet · Cross super-category vehicle | PCK: 28.35 -> 34.92 | Improved |
Closest papers by meaning
- CapeX: Category-Agnostic Pose Estimation from Textual Point Explanation (ICLR 2025 · similarity 0.92)
- Meta-Point Learning and Refining for Category-Agnostic Pose Estimation (CVPR 2024 · similarity 0.90)
- ESCAPE: Encoding Super-keypoints for Category-Agnostic Pose Estimation (CVPR 2024 · similarity 0.90)
- Rethinking Correspondence-based Category-Level Object Pose Estimation (CVPR 2025 · similarity 0.89)
- Universal Features Guided Zero-Shot Category-Level Object Pose Estimation (AAAI 2025 · similarity 0.88)
- Instance-Adaptive and Geometric-Aware Keypoint Learning for Category-Level 6D Object Pose Estimation (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.
