Research map / 3D vision / Human pose estimation
Cross-View Tracking for Multi-Human 3D Pose Estimation at Over 100 FPS
Check 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 |
|---|---|---|---|
| use cross-view tracking with matching directly between 2D detections and retained 3D poses | multi-human 3D pose estimation · Campus | PCP average (%): 96.3 -> 96.6 | Helped |
| use cross-view tracking with matching directly between 2D detections and retained 3D poses | multi-human 3D pose estimation · Shelf | FPS: 9.5 -> 325 | Helped |
| remove normalization from incremental triangulation (ablation) | incremental 3D reconstruction · Campus | PCP whole: 96.6 -> 95.4 | Hurt |
| remove normalization from incremental triangulation (ablation) | incremental 3D reconstruction · Shelf | PCP whole: 96.8 -> 96.9 | No clear effect |
| replace incremental triangulation with conventional triangulation (ablation) | multi-human 3D pose estimation · Campus | PCP whole: 96.6 -> 94.4 | Hurt |
| replace incremental triangulation with conventional triangulation (ablation) | multi-human 3D pose estimation · Shelf | PCP whole: 96.8 -> 94.8 | Hurt |
| use only 2D correspondence and remove 3D correspondence (ablation) | cross-view tracking affinity · Campus | Association Accuracy (%): 96.38 -> 45.69 | Hurt |
| use only 2D correspondence and remove 3D correspondence (ablation) | cross-view tracking affinity · Campus | PCP (%): 96.63 -> 62.29 | Hurt |
| use only 3D correspondence and remove 2D correspondence (ablation) | cross-view tracking affinity · Campus | Association Accuracy (%): 96.38 -> 96.38 | No clear effect |
| use only 3D correspondence and remove 2D correspondence (ablation) | cross-view tracking affinity · Campus | PCP (%): 96.63 -> 96.49 | No clear effect |
| add 3D correspondence to 2D-only affinity (ablation) | cross-view tracking affinity · Campus | Association Accuracy (%): 45.69 -> 96.22 | Helped |
| match in 3-space rather than pairwise in camera coordinates (ablation) | cross-view human association · Campus Camera3 | association accuracy: 32%, 5.2%, 9.2%, 4.6% | Helped |
Closest papers by meaning
- Fast and Robust Multi-Person 3D Pose Estimation From Multiple Views (CVPR 2019 · similarity 0.91)
- Multi-View Multi-Person 3D Pose Estimation With Plane Sweep Stereo (CVPR 2021 · similarity 0.90)
- CoMotion: Concurrent Multi-person 3D Motion (ICLR 2025 · similarity 0.88)
- TrackingWorld: World-centric Monocular 3D Tracking of Almost All Pixels (NeurIPS 2025 · similarity 0.87)
- Optimizing Human Pose Estimation Through Focused Human and Joint Regions (AAAI 2025 · similarity 0.87)
- Motions as Queries: One-Stage Multi-Person Holistic Human Motion Capture (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.