Research map / 3D vision / Human pose estimation
Towards Alleviating the Modeling Ambiguity of Unsupervised Monocular 3D Human Pose Estimation
Zhenbo Yu, Bingbing Ni, Jingwei Xu, Junjie Wang, Chenglong Zhao, Wenjun Zhang
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 multi-view motion consistency loss Ltmc (ablation) | lifting baseline · Human3.6M (GT 2D) | PMPJPE: 46.0 -> 43.5 | Improved |
| add multi-view motion consistency loss Ltmc (ablation) | lifting baseline · Human3.6M (GT 2D) | MPJPE: 105.0 -> 101.7 | Improved |
| add KL scale distribution loss Lkl (ablation) | lifting+Ltmc · Human3.6M (GT 2D) | MPJPE: 101.7 -> 96.0 | Improved |
| add KL scale distribution loss Lkl (ablation) | lifting+Ltmc · Human3.6M (GT 2D) | PMPJPE: 43.5 -> 42.9 | Improved |
| add bone consistency loss Lbone (ablation) | lifting+Ltmc+Lkl · Human3.6M (GT 2D) | MPJPE: 96.0 -> 85.3 | Improved |
| add bone consistency loss Lbone (ablation) | lifting+Ltmc+Lkl · Human3.6M (GT 2D) | PMPJPE: 42.9 -> 42.0 | Improved |
| add scale estimation module on 2D pose (ablation) | lifting with universal scale · Human3.6M | MPJPE: 105.0 -> 85.3 | Improved |
| add scale estimation module on 2D pose (ablation) | lifting with no normalization · Human3.6M | MPJPE: 260.7 -> 233.4 | Improved |
| use universal scale normalization instead of step-wise (ablation) | lifting · Human3.6M | PMPJPE: 48.5 -> 46.0 | Improved |
| apply scale module on 2D input rather than 3D output (ablation) | lifting · Human3.6M | MPJPE: 97.5 -> 85.3 | Improved |
| apply scale module on 2D input rather than 3D output (ablation) | lifting · Human3.6M | PMPJPE: 43.6 -> 42.0 | Improved |
| feed reference scale from ground truth instead of learned scale | lifting · Human3.6M | PMPJPE: 42.0 -> 39.7 | Improved |
Closest papers by meaning
- Robust 3D Human Pose Estimation from Single Images or Video Sequences (TPAMI 2018 · similarity 0.90)
- CoDePose: Multi-View 3D Human Pose Estimation via Coupled 2D-3D Denoising Diffusion (ECCV 2026 · similarity 0.89)
- ManiPose: Manifold-Constrained Multi-Hypothesis 3D Human Pose Estimation (NeurIPS 2024 · similarity 0.89)
- Lifting by Image – Leveraging Image Cues for Accurate 3D Human Pose Estimation (AAAI 2024 · similarity 0.88)
- PoseIRM: Enhance 3D Human Pose Estimation on Unseen Camera Settings via Invariant Risk Minimization (CVPR 2024 · similarity 0.87)
- PandaPose: 3D Human Pose Lifting from a Single Image via Propagating 2D Pose Prior to 3D Anchor Space (NeurIPS 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.
