Research map / 3D vision / Human pose estimation
TesseTrack: End-to-End Learnable Multi-Person Articulated 3D Pose Tracking
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 backbone features instead of 2D joint heatmaps (ablation) | TesseTrack · Panoptic | MPJPE (mm): 16.3 -> 13.8 | Helped |
| add temporal tesseract prediction (ablation) | TesseTrack · Panoptic | MPJPE (mm): 13.8 -> 8.0 | Helped |
| add learned descriptor merging instead of heatmap averaging (ablation) | TesseTrack with heuristic matching · Panoptic | MPJPE (mm): 8.1 -> 7.5 | Helped |
| replace heuristic Hungarian matching with differentiable learned matching (ablation) | TesseTrack with learned descriptor merging · Panoptic | MPJPE (mm): 7.5 -> 7.3 | Helped |
| add temporal tesseract prediction (ablation) | TesseTrack · Human3.6M, multi-view | MPJPE (mm), Total: 20.4 -> 18.7 | Helped |
| use temporal person descriptors rather than instantaneous descriptors (ablation) | TesseTrack · Panoptic | 3D MOTA, Avg: 87.6 -> 92.1 | Helped |
| replace Hungarian matching with a learned end-to-end matcher (ablation) | TesseTrack with temporal descriptors · Panoptic | 3D MOTA, Avg: 92.1 -> 94.1 | Helped |
| use TesseTrack instead of Tu et al. | multi-person 3D pose estimation · Panoptic, 5 views | MPJPE (mm): 17.7 -> 7.3 | Helped |
| use TesseTrack instead of Tu et al. | multi-person 3D pose estimation · Panoptic, monocular | MPJPE (mm): 51.1 -> 18.9 | Helped |
| use TesseTrack instead of Tu et al. | multi-person 3D pose estimation · TUM Campus | 3D-PCK, Total: 96.7 -> 97.4 | Helped |
| use TesseTrack instead of Tu et al. | multi-person 3D pose estimation · TUM Shelf | 3D-PCK, Total: 97.0 -> 98.2 | Helped |
| use TesseTrack instead of Iskakov et al. | single-person 3D pose estimation · Human3.6M, Protocol #1, multi-view | MPJPE (mm), Total: 20.8 -> 18.7 | Helped |
Closest papers by meaning
- ArtTrack: Articulated Multi-Person Tracking in the Wild (CVPR 2017 · similarity 0.91)
- PoseTrack: Joint Multi-Person Pose Estimation and Tracking (CVPR 2017 · similarity 0.90)
- Direct Multi-view Multi-person 3D Pose Estimation (NeurIPS 2021 · similarity 0.90)
- CoMotion: Concurrent Multi-person 3D Motion (ICLR 2025 · similarity 0.89)
- SelfPose3d: Self-Supervised Multi-Person Multi-View 3d Pose Estimation (CVPR 2024 · similarity 0.87)
- FreeMan: Towards Benchmarking 3D Human Pose Estimation under Real-World Conditions (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.