Research map / 3D vision / Human pose estimation
3D Human Pose Estimation = 2D Pose Estimation + Matching
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 |
|---|---|---|---|
| predict 2D pose, then match and warp a 3D exemplar | 3D human pose estimation · Human3.6M Protocol 1 | MPJPE: 82.72 vs 108.3; 88.1 | Helped |
| use 2D pose estimation plus matching without temporal input | 3D human pose estimation · Human3.6M Protocol 2 | MPJPE: 114.18 vs 113.01; 124.97 | Mixed |
| add exemplar warping (ablation) | nearest-neighbor 3D pose matching · Human3.6M Protocol 1, predicted 2D | MPJPE: 85.52 -> 82.72 | Helped |
| add exemplar warping (ablation) | nearest-neighbor 3D pose matching · Human3.6M Protocol 1, ground-truth 2D | MPJPE: 70.93 -> 57.50 | Helped |
| use the optimal 3D exemplar and warp it (ablation) | 3D exemplar matching · Human3.6M Protocol 1, optimal exemplar match | MPJPE: 60.11 -> 37.32 | Helped |
| increase the library from 18k to 180k pairs (ablation) | 3D exemplar matching · Human3.6M Protocol 1 | MPJPE: 85.94 -> 82.37 | Helped |
| increase the library from 180k to 1,800k pairs (ablation) | 3D exemplar matching · Human3.6M Protocol 1 | MPJPE: 82.37 -> 82.72 | Hurt |
| add exemplar warping (ablation) | Human3.6M-trained exemplar matching · HumanEva-I, ground-truth 2D | MPJPE: 92.17 -> 68.29 | Helped |
| increase the shortlist from k=1 to k=10 and search over virtual cameras (ablation) | warped exemplar matching · Human3.6M Protocol 2, ground-truth 2D | MPJPE: 51.06 -> 49.55 | Helped |
Closest papers by meaning
- Lifting From the Deep: Convolutional 3D Pose Estimation From a Single Image (CVPR 2017 · similarity 0.94)
- Coarse-To-Fine Volumetric Prediction for Single-Image 3D Human Pose (CVPR 2017 · similarity 0.92)
- 3D Human Pose Estimation From a Single Image via Distance Matrix Regression (CVPR 2017 · similarity 0.92)
- MoCap-guided Data Augmentation for 3D Pose Estimation in the Wild (NeurIPS 2016 · similarity 0.91)
- Learning to Estimate 3D Human Pose and Shape From a Single Color Image (CVPR 2018 · similarity 0.90)
- PandaPose: 3D Human Pose Lifting from a Single Image via Propagating 2D Pose Prior to 3D Anchor Space (NeurIPS 2025 · similarity 0.89)
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.