Research map / 3D vision / Human pose estimation
Differentiable Dynamics for Articulated 3D Human Motion Reconstruction
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 |
|---|---|---|---|
| add differentiable physics-based reconstruction | Kinematic estimates · Human3.6M | MPJPE-G: 145.3 -> 139.1 | Helped |
| add differentiable physics-based reconstruction | Kinematic estimates · Human3.6M | Foot skate (%): 47.5 -> 7.4 | Helped |
| add differentiable physics-based reconstruction | Kinematic estimates · AIST | MPJPE-G: 155.7 -> 150.2 | Helped |
| add differentiable physics-based reconstruction | Kinematic estimates · AIST | MPJPE-2d: 10.4 -> 12.1 | Hurt |
| add differentiable physics-based reconstruction | Kinematic estimates · AIST | Foot skate (%): 50.9 -> 19.6 | Helped |
| use Basin-Hopping with BFGS instead of local BFGS alone | DiffPhy optimization · Human3.6M validation | MPJPE-G: 160.1 -> 144.9 | Helped |
| use BFGS instead of CMA-ES | DiffPhy optimization · Human3.6M validation | MPJPE-G: 206.7 -> 160.1 | Helped |
| add root position loss (ablation) | DiffPhy · Human3.6M ablation subset | MPJPE-G: 165.7 -> 144.9 | Helped |
| add 2d keypoint loss (ablation) | DiffPhy · Human3.6M ablation subset | MPJPE-2d: 17.1 -> 12.6 | Helped |
| add joint angle loss (ablation) | DiffPhy · Human3.6M ablation subset | MPJPE-G: 156.8 -> 144.9 | Helped |
| add root position and joint angle losses (ablation) | DiffPhy · Human3.6M ablation subset | MPJPE-G: 216.6 -> 144.9 | Helped |
| add anatomical joint limits (ablation) | DiffPhy · Human3.6M ablation subset | MPJPE-PA: 62.2 -> 61.1 | Helped |
Closest papers by meaning
- MultiPhys: Multi-Person Physics-aware 3D Motion Estimation (CVPR 2024 · similarity 0.88)
- PhysPT: Physics-aware Pretrained Transformer for Estimating Human Dynamics from Monocular Videos (CVPR 2024 · similarity 0.87)
- D^3-Human: Dynamic Disentangled Digital Human from Monocular Video (CVPR 2025 · similarity 0.87)
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.87)
- SurMo: Surface-based 4D Motion Modeling for Dynamic Human Rendering (CVPR 2024 · similarity 0.87)
- UniCon3R: Unified Contact-aware 4D Human-Scene Reconstruction from Monocular Video (NeurIPS 2026 · 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.