Research map / 3D vision / Human pose estimation
MetricHMSR: Metric Human Mesh and Scene Recovery from Monocular Images
Chentao Song, He Zhang, Haolei Yuan, Haozhe Lin, Jianhua Tao, Hongwen Zhang, Tao Yu
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 bounding ray map input (ablation) | MetricHMR image-only · EMDB-2 | W-MPJPE: 191.8 -> 154.1 | Improved |
| add bounding ray map input (ablation) | MetricHMR image-only · 3DPW | PA-MPJPE: 35.6 -> 34.4 | Improved |
| add HumanMoE on top of ray map (ablation) | MetricHMR · 3DPW | PA-MPJPE: 34.4 -> 33.6 | Improved |
| add HumanMoE on top of ray map (ablation) | MetricHMR · EMDB-2 | W-MPJPE: 154.1 -> 152.5 | Improved |
| add Patch MoE to Global MoE (ablation) | HumanMoE · 3DPW | PA-MPJPE: 34.1 -> 33.6 | Improved |
| add Global MoE to Patch MoE (ablation) | HumanMoE · 3DPW | PA-MPJPE: 34.2 -> 33.6 | Improved |
| use 4 routed experts instead of 0 (ablation) | HumanMoE · EMDB-2 | WA-MPJPE: 60.8 -> 55.6 | Improved |
| increase routed experts from 4 to 8 (ablation) | HumanMoE · 3DPW | PA-MPJPE: 33.6 -> 33.9 | No clear effect |
| increase routed experts from 4 to 32 (ablation) | HumanMoE · EMDB-2 | WA-MPJPE: 55.6 -> 56.3 | Hurt |
| human-guided metric depth refinement | MapAnything · PROX | AbsRel: 0.18 -> 0.13 | Improved |
| MetricHMSR vs PromptHMR | local pose · EMDB-1 | PA-MPJPE: 40.1 -> 43.2 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- Human3R — Human3R: Everyone Everywhere All at Once (ICLR 2026) · better on 3 of 3 shared results
- TRAM — TRAM: Global Trajectory and Motion of 3D Humans from in-the-wild Videos (ECCV 2024) · better on 3 of 3 shared results
- PromptHMR — PromptHMR: Promptable Human Mesh Recovery (CVPR 2025) · better on 2 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
- Metric from Human: Zero-shot Monocular Metric Depth Estimation via Test-time Adaptation (NeurIPS 2024 · similarity 0.89)
- UniDepth: Universal Monocular Metric Depth Estimation (CVPR 2024 · similarity 0.89)
- Synergistic Global-space Camera and Human Reconstruction from Videos (CVPR 2024 · similarity 0.88)
- Reconstructing People, Places, and Cameras (CVPR 2025 · similarity 0.88)
- ODHSR: Online Dense 3D Reconstruction of Humans and Scenes from Monocular Videos (CVPR 2025 · similarity 0.87)
- BLADE: Single-view Body Mesh Estimation through Accurate Depth Estimation (CVPR 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.
