Research map / 3D vision / Human pose estimation
Towards Robust and Expressive Whole-body Human Pose and Shape Estimation
Hui En Pang, Zhongang Cai, Lei Yang, Qingyi Tao, Zhonghua Wu, Tianwei Zhang, Ziwei Liu
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 all location features (ablation) | Base (DR54) hand model · FreiHAND | PA-PVE: 7.67 -> 7.46 | Improved |
| add strong augmentation (ablation) | HMR hand model · FreiHAND | PA-PVE: 7.85 -> 8.11 | Hurt |
| add contrastive loss and positive samples (ablation) | LF (all) hand model · FreiHAND | PVE: 15.59 -> 14.57 | Improved |
| add projected segmentation loss (ablation) | LF (all) hand model with part segmentation supervision · FreiHAND | MPJPE: 14.92 -> 14.38 | Improved |
| add all location features (ablation) | HMR body model · 3DPW test | MPJPE: 96.2 -> 105.7 | Hurt |
| use keypoints rather than pose rotations as the contrastive representation, with positive samples (ablation) | RoboSMPLX body model · 3DPW test | PA-MPJPE: 55.9 -> 55.4 | Improved |
| use pose plus global orientation as the contrastive representation (ablation) | RoboSMPLX body model · 3DPW test | PA-MPJPE: 55.9 -> 76.8 | Hurt |
| use pose-variant augmentation to form positive samples (ablation) | contrastive hand model · FreiHAND | PA-PVE: 7.67 -> 8.15 | Hurt |
| use color and location augmentations for positive samples (ablation) | contrastive hand model · FreiHAND | PVE: 16.08 -> 14.77 | Improved |
| add contrastive learning to strong augmentation (ablation) | HMR hand model · FreiHAND | PA-PVE: 8.11 -> 7.41 | Improved |
| use RoboSMPLX instead of Hand4Whole | whole-body hand estimation · FreiHAND positional augmentations | PVE under Transy +0.2y: 22.51 -> 15.50 | Improved |
| use RoboSMPLX instead of ExPose | whole-body face estimation · EHF positional augmentations | 3DRMSE under Transy +0.2y: 2.46 -> 2.13 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- PyMAF * — PyMAF: 3D Human Pose and Shape Regression With Pyramidal Mesh Alignment Feedback Loop (ICCV 2021) · better on 0 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Single-Network Whole-Body Pose Estimation (ICCV 2019 · similarity 0.90)
- AlphaPose: Whole-Body Regional Multi-Person Pose Estimation and Tracking in Real-Time (TPAMI 2022 · similarity 0.89)
- AiOS: All-in-One-Stage Expressive Human Pose and Shape Estimation (CVPR 2024 · similarity 0.89)
- Probabilistic 3D Human Shape and Pose Estimation From Multiple Unconstrained Images in the Wild (CVPR 2021 · similarity 0.88)
- ZoomNAS: Searching for Whole-body Human Pose Estimation in the Wild (TPAMI 2022 · similarity 0.88)
- SMPLest-X: Ultimate Scaling for Expressive Human Pose and Shape Estimation (TPAMI 2025 · similarity 0.88)
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.
