Research map / Point clouds / Shape and surface completion
Skeleton Merger: An Unsupervised Aligned Keypoint Detector
Ruoxi Shi, Zhengrong Xue, Yang You, 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 Skeleton Merger instead of Fernandez et al. | Unsupervised keypoint detection · KeypointNet Airplane | DAS: 61.4 -> 77.7 | Improved |
| Use Skeleton Merger instead of Fernandez et al. | Unsupervised keypoint detection · KeypointNet Chair | DAS: 64.3 -> 76.8 | Improved |
| Use Skeleton Merger instead of Fernandez et al. | Unsupervised keypoint detection · SyncSpecCNN Chair | DAS: 54.2 -> 73.8 | Improved |
| Add activation strengths for masking nonexistent skeleton edges (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | mIoU: 55.5 -> 79.4; 8.4 -> 68.4 | Improved |
| Add activation strengths for masking nonexistent skeleton edges (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | DAS: 72.1 -> 77.7; 65.2 -> 76.8 | Improved |
| Add fidelity loss to Composite Chamfer Distance (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | mIoU: 78.2 -> 79.4; 60.0 -> 68.4 | Improved |
| Add fidelity loss to Composite Chamfer Distance (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | DAS: 76.2 -> 77.7; 74.4 -> 76.8 | Improved |
| Add coverage loss to Composite Chamfer Distance (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | mIoU: 17.8 -> 79.4; 1.1 -> 68.4 | Improved |
| Add coverage loss to Composite Chamfer Distance (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | DAS: 35.6 -> 77.7; 37.3 -> 76.8 | Improved |
| Remove learned pointwise offsets from skeleton edges (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | mIoU: 79.4 -> 85.6; 68.4 -> 62.0 | Mixed |
| Add learned pointwise offsets to skeleton edges (ablation) | Skeleton Merger · KeypointNet Airplane and Chair | DAS: 72.4 -> 77.7; 75.6 -> 76.8 | Improved |
Later papers that compared with this method
2 later accepted papers report SkeletonMerger as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- KeypointDETR: An End-to-End 3D Keypoint Detector (ECCV 2024 · better on 1 of 1 shared result)
- Unsupervised Learning of 3D Semantic Keypoints with Mutual Reconstruction (ECCV 2022 · better on 5 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.
- PointConv — PointConv: Deep Convolutional Networks on 3D Point Clouds (CVPR 2019) · better on 0 of 3 shared results
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
- Key-Grid: Unsupervised 3D Keypoints Detection using Grid Heatmap Features (NeurIPS 2024 · similarity 0.89)
- KeypointDETR: An End-to-End 3D Keypoint Detector (ECCV 2024 · similarity 0.87)
- KeypointDeformer: Unsupervised 3D Keypoint Discovery for Shape Control (CVPR 2021 · similarity 0.87)
- Unsupervised 3D Structure Inference from Category-Specific Image Collections (CVPR 2024 · similarity 0.87)
- KeyPose: Category-Level 6D Object Pose Estimation with Self-Adaptive Keypoints (AAAI 2025 · similarity 0.85)
- SPEAL: Skeletal Prior Embedded Attention Learning for Cross-Source Point Cloud Registration (AAAI 2024 · 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.
