Research map / 3D vision / Human pose estimation
MoCap-guided Data Augmentation for 3D Pose Estimation in the Wild
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 |
|---|---|---|---|
| train classifier on synthetic + real images rather than real only (190k) (ablation) | K-way pose classifier (AlexNet) · Human3.6M P1 | 3D error (mm): 97.7 -> 88.1 | Helped |
| use synthetic images instead of real images for the classifier (same 190k poses) (ablation) | K-way pose classifier · Human3.6M P1 | 3D error (mm): 97.2 vs 97.7 | No clear effect |
| use classifier instead of regressor at small data (17k) (ablation) | pose estimation head · Human3.6M P1 | 3D error (mm): 112.9 (Reg.) vs 149.7 (Class.) | Hurt |
| use in-the-wild MPII+LSP 2D source instead of H3.6M images (ablation) | synthetic data generation · LSP 2D error | 2D error (pix): 31.1 -> 20.7 | Helped |
| use MPII+LSP 2D source instead of H3.6M images (ablation) | synthetic data generation · H3.6M 2D error | 2D pose error (pix): 8.8 -> 17.3 | Hurt |
| scale CMU synthetic set from 190k to 2.1M poses (ablation) | synthetic data generation · LSP and H3.6M 2D error | 2D pose error (pix): LSP 22.4 -> 13.8; H3.6M 19.7 -> 11.2 | Helped |
| fine-tune VGG-16 instead of AlexNet for pose classification | K-way pose classifier · LSP | 2D pose error (pix): 13.8 -> 11.5 | Helped |
| retrain AlexNet from scratch instead of ImageNet fine-tuning (ablation) | AlexNet pose classifier · Human3.6M P1 | 3D error (mm): 98.3 -> 88.1 | Helped |
| use CMU 3D poses instead of H3.6M 3D poses (190k) (ablation) | synthetic data generation · LSP 2D error | 2D pose error (pix): 20.7 -> 22.4 | Hurt |
Closest papers by meaning
- 3D Human Pose Estimation = 2D Pose Estimation + Matching (CVPR 2017 · similarity 0.91)
- In the Wild Human Pose Estimation Using Explicit 2D Features and Intermediate 3D Representations (CVPR 2019 · similarity 0.90)
- Sim2real transfer learning for 3D human pose estimation: motion to the rescue (NeurIPS 2019 · similarity 0.89)
- PoseGen: Learning to Generate 3D Human Pose Dataset with NeRF (AAAI 2024 · similarity 0.87)
- Lifting by Image – Leveraging Image Cues for Accurate 3D Human Pose Estimation (AAAI 2024 · similarity 0.87)
- Neural Localizer Fields for Continuous 3D Human Pose and Shape Estimation (NeurIPS 2024 · similarity 0.86)
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.