Research map / 3D vision / Human pose estimation
Discovery of Latent 3D Keypoints via End-to-end Geometric Reasoning
Read the paperCheck 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 |
|---|---|---|---|
| Replace annotated keypoint supervision with end-to-end pose and multi-view consistency objectives | KeypointNet versus supervised keypoint baseline · ShapeNet cars test | mean angular distance error: 13.961 -> 11.310 | Helped |
| Replace annotated keypoint supervision with end-to-end pose and multi-view consistency objectives | KeypointNet versus supervised keypoint baseline · ShapeNet chairs test | mean angular distance error: 20.502 -> 14.572 | Helped |
| Replace annotated keypoint supervision with end-to-end pose and multi-view consistency objectives | KeypointNet versus supervised keypoint baseline · ShapeNet planes test | mean angular distance error: 17.800 -> 17.330 | Helped |
| Add a predicted orientation flag | Supervised keypoint baseline · ShapeNet cars, planes, and chairs test | mean angular distance error: 16.268 -> 13.961; 18.350 -> 17.800; 21.882 -> 20.502 | Helped |
| Remove the multi-view consistency loss (ablation) | KeypointNet · ShapeNet keypoint tracking across views | keypoint consistency | Hurt |
| Remove the pose estimation loss (ablation) | KeypointNet · ShapeNet keypoint selection | keypoint spatial distribution | Hurt |
| Increase noise in the pose estimation loss (ablation) | KeypointNet · ShapeNet keypoint selection | keypoint spatial distribution | Helped |
| Remove silhouette consistency (ablation) | KeypointNet · ShapeNet keypoint locations | silhouette adherence | Hurt |
| Vary the number of predicted keypoints (ablation) | KeypointNet · ShapeNet object categories | coverage of object parts | Helped |
Closest papers by meaning
- KeypointNet: A Large-Scale 3D Keypoint Dataset Aggregated From Numerous Human Annotations (CVPR 2020 · similarity 0.90)
- Back to 3D: Few-Shot 3D Keypoint Detection with Back-Projected 2D Features (CVPR 2024 · similarity 0.89)
- Unsupervised 3D Structure Inference from Category-Specific Image Collections (CVPR 2024 · similarity 0.89)
- KeyPose: Category-Level 6D Object Pose Estimation with Self-Adaptive Keypoints (AAAI 2025 · similarity 0.88)
- Key-Grid: Unsupervised 3D Keypoints Detection using Grid Heatmap Features (NeurIPS 2024 · similarity 0.88)
- Structure-Aware Correspondence Learning for Relative Pose Estimation (CVPR 2025 · 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.