Research map / 3D vision / Human pose estimation
HuMMan: Multi-modal 4D Human Dataset for Versatile Sensing and Modeling
Zhongang Cai, Daxuan Ren, Ailing Zeng, Zhengyu Lin, Tao Yu, Wenjia Wang, Xiangyu Fan, Yang Gao, Yifan Yu, Liang Pan, Fangzhou Hong, Mingyuan Zhang and 3 more
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 |
|---|---|---|---|
| train on HuMMan instead of H36M (ablation) | FCN · 3D keypoint test on AIST++ | MPJPE (mm): 133.9 -> 116.4 | Improved |
| train on HuMMan instead of H36M (ablation) | Video3D · 3D keypoint test on AIST++ | MPJPE (mm): 128.5 -> 109.2 | Improved |
| train on iPhone data instead of Kinect data (ablation) | HMR · HuMMan iPhone test set | MPJPE: 97.81 -> 72.62 | Improved |
| train on iPhone data instead of Kinect data (ablation) | VoteHMR · HuMMan iPhone test set | MPJPE: 255.71 -> 83.18 | Improved |
| use point cloud input (VoteHMR) instead of image input (HMR) (ablation) | HMR · HuMMan parametric human recovery | MPJPE: 54.78 -> 144.99 | Hurt |
| use multi-frame Video3D instead of single-frame FCN (ablation) | FCN · HuMMan 3D keypoint (in-domain) | MPJPE (mm): 78.5 -> 73.1 | Improved |
| use Function4D instead of PIFu for geometry (ablation) | PIFu · HuMMan geometry reconstruction | Chamfer distance (1e-2 m): 7.92 -> 1.80 | Improved |
| train on one view only and test on opposite view (View 5 vs View 0) (ablation) | FCN · HuMMan cross-view (P3) | MPJPE (mm): 66.4 -> 268.4 | Hurt |
| train on one view only and test on opposite view (View 5 vs View 0) (ablation) | HMR · HuMMan cross-view (P3) | MPJPE: 61.9 -> 421.0 | Hurt |
| test on unseen action category, whole body (P2) rather than same-distribution split (ablation) | FCN · HuMMan cross-action (P2) | MPJPE (mm): 78.5 -> 97.5 | Hurt |
Closest papers by meaning
- GeneMAN: Generalizable Single-Image 3D Human Reconstruction from Multi-Source Human Data (NeurIPS 2025 · similarity 0.84)
- Reconstructing People, Places, and Cameras (CVPR 2025 · similarity 0.84)
- Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives (CVPR 2024 · similarity 0.84)
- FreeMan: Towards Benchmarking 3D Human Pose Estimation under Real-World Conditions (CVPR 2024 · similarity 0.83)
- Harmony4D: A Video Dataset for In-The-Wild Close Human Interactions (NeurIPS 2024 · similarity 0.83)
- RELI11D: A Comprehensive Multimodal Human Motion Dataset and Method (CVPR 2024 · similarity 0.83)
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.
