Research map / 3D vision / Human pose estimation
Delving Deep into Pixel Alignment Feature for Accurate Multi-View Human Mesh Recovery
Read the paperCheck 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 |
|---|---|---|---|
| Use PaFF with calibrated cameras instead of the previous multi-view method | Multi-view human mesh recovery · Human3.6M | MPJPE: 46.9 -> 33.0 | Helped |
| Use max-pooling rather than average pooling for pose and shape fusion (ablation) | PaFF · Human3.6M | MPJPE: 36.1 -> 33.0 | Helped |
| Use max-pooling rather than softmax-then-sum for pose and shape fusion (ablation) | PaFF · Human3.6M | MPJPE: 34.9 -> 33.0 | Helped |
| Replace transformer-plus-max-pooling with max-pooling alone for pose and shape fusion (ablation) | PaFF · Human3.6M | MPJPE: 34.4 -> 33.0 | Helped |
| Add Global O Aligner using camera rotations (ablation) | Independent orientation estimation in PaFF · Human3.6M | MPJPE: 44.9 -> 34.3 | Helped |
| Add transformer-based cross-view orientation fusion to aligned independent orientation estimates (ablation) | PaFF orientation estimation · Human3.6M | MPJPE: 34.3 -> 33.0 | Helped |
| Add transformer-based cross-view orientation fusion to aligned independent orientation estimates (ablation) | PaFF orientation estimation · Human3.6M | O Err: $5.8^{\circ}$ -> $\textbf{5.1}^{\circ}$ | Helped |
| Estimate global translation and adaptive body scale (ablation) | PaFF · Human3.6M | MPJPE: 36.5 -> 33.0 | Helped |
| Estimate global translation and adaptive body scale (ablation) | PaFF · Human3.6M | PVE: 58.6 -> 48.9 | Helped |
| Estimate global translation and adaptive body scale (ablation) | PaFF · Human3.6M | PA-MPJPE: 26.9 -> 26.9 | No clear effect |
| Use calibration-free rather than calibrated cameras | PaFF · Human3.6M | MPJPE: 33.0 -> 44.8 | Hurt |
Closest papers by meaning
- FactorizedHMR: A Hybrid Framework for Video Human Mesh Recovery (NeurIPS 2026 · similarity 0.87)
- ADHMR: Aligning Diffusion-based Human Mesh Recovery via Direct Preference Optimization (ICML 2025 · similarity 0.87)
- BLADE: Single-view Body Mesh Estimation through Accurate Depth Estimation (CVPR 2025 · similarity 0.86)
- Fine Structure-Aware Sampling: A New Sampling Training Scheme for Pixel-Aligned Implicit Models in Single-View Human Reconstruction (AAAI 2024 · similarity 0.86)
- Incorporating Test-Time Optimization into Training with Dual Networks for Human Mesh Recovery (NeurIPS 2024 · similarity 0.85)
- GenHMR: Generative Human Mesh Recovery (AAAI 2025 · similarity 0.85)
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.