Research map / 3D vision / Human pose estimation
Multi-person 3D Pose Estimation in Crowded Scenes Based on Multi-View Geometry
He Chen, Pengfei Guo, Pengfei Li, Gim Hee Lee, Gregory Chirikjian
Read the paperCheck 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 |
|---|---|---|---|
| add MLE as an intermediate step before MAP (ablation) | MAP triangulation · WildTrack | average reprojection error: 64.75 -> 38.55 | Improved |
| add MLE as an intermediate step before MAP (ablation) | MAP triangulation · WildTrack | maximum reprojection error: 316.7 -> 219.29 | Improved |
| use proposed MAP instead of vanilla triangulation (ablation) | multi-view 3D pose · Chariot Mk I | AP: 60.0 -> 89.8 | Improved |
| use proposed optimization instead of vanilla triangulation (ablation) | multi-view 3D pose · LOEWENPLATZ | AP: 66.7 -> 81.8 | Improved |
| use proposed MAP instead of vanilla triangulation (ablation) | multi-view 3D pose · Wildtrack | AP: 55.3 -> 70.0 | Improved |
| use proposed MAP instead of vanilla triangulation (ablation) | multi-view 3D pose · Wildtrack | AR: 77.3 -> 78.3 | No clear effect |
| feet-based graph matching instead of epipolar constraint | correspondence · CMU Panoptic | precision: 86.5 -> 93.7 | Improved |
| feet-based graph matching instead of RANSAC | correspondence · CMU Panoptic | precision: 46.0 -> 93.7 | Improved |
| add extra attention loss on foot joints (modified SPPE) | 2D pose network · COCO foot val | AP: 77.9 -> 80.1 | Improved |
| add extra attention loss on foot joints (modified SPPE) | 2D pose network · COCO foot val | AR: 82.5 -> 82.0 | No clear effect |
| our SPPE vs Cao et al. on body+foot | 2D pose network · COCO body+foot val | AP: 65.3 -> 65.3 | No clear effect |
Closest papers by meaning
- Fast and Robust Multi-Person 3D Pose Estimation From Multiple Views (CVPR 2019 · similarity 0.90)
- CrowdPose: Efficient Crowded Scenes Pose Estimation and a New Benchmark (CVPR 2019 · similarity 0.89)
- Shape-Aware Multi-Person Pose Estimation From Multi-View Images (ICCV 2021 · similarity 0.89)
- Fast and Robust Multi-Person 3D Pose Estimation and Tracking From Multiple Views (TPAMI 2021 · similarity 0.89)
- Dual Networks Based 3D Multi-Person Pose Estimation From Monocular Video (TPAMI 2022 · similarity 0.88)
- COMPOSE: Hypergraph Cover Optimization for Multi-view 3D Human Pose Estimation (NeurIPS 2026 · 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.
