Research map / 3D vision / Human pose estimation
End-to-End Human Pose and Mesh Reconstruction with Transformers
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 |
|---|---|---|---|
| Use METRO instead of VIBE | 3DPW single-image human mesh reconstruction · 3DPW test | MPVE: 99.1 -> 88.2 | Helped |
| Use METRO instead of VIBE | 3DPW single-image human pose reconstruction · 3DPW test | MPJPE: 82.0 -> 77.1 | Helped |
| Use METRO instead of VIBE | 3DPW human mesh reconstruction · 3DPW test | PA-MPJPE: 51.9 -> 47.9 | Helped |
| Use METRO instead of the strongest listed baseline | Human3.6M human pose reconstruction · Human3.6M test | MPJPE: 55.7 -> 54.0 | Helped |
| Use METRO instead of the strongest listed baseline | Human3.6M human pose reconstruction · Human3.6M test | PA-MPJPE: 39.3 -> 36.7 | Helped |
| Use METRO instead of I2LMeshNet | FreiHAND mesh reconstruction · FreiHAND test | PA-MPVPE: 7.6 -> 6.3 | Helped |
| Use METRO instead of I2LMeshNet | FreiHAND pose reconstruction · FreiHAND test | PA-MPJPE: 7.4 -> 6.5 | Helped |
| Add Masked Vertex Modeling (ablation) | METRO · Human3.6M | PA-MPJPE: 39.1 -> 36.7 | Helped |
| Increase maximum masked-query percentage from 0% to 30% (ablation) | METRO · Human3.6M | PA-MPJPE: 39.1 -> 36.7 | Helped |
| Increase maximum masked-query percentage from 30% to 40% (ablation) | METRO · Human3.6M | PA-MPJPE: 36.7 -> 38.2 | Hurt |
| Use HRNet-W64 instead of ResNet50 (ablation) | METRO · Human3.6M | PA-MPJPE: 40.6 -> 36.7 | Helped |
| Use progressive dimensionality reduction instead of direct projection (ablation) | Transformer mesh regression · Human3.6M validation | PA-MPJPE: 208.7 -> 36.7 | Helped |
Closest papers by meaning
- Reconstructing Hands in 3D with Transformers (CVPR 2024 · similarity 0.86)
- End-to-End HOI Reconstruction Transformer with Graph-based Encoding (CVPR 2025 · similarity 0.86)
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.86)
- MEGA: Masked Generative Autoencoder for Human Mesh Recovery (CVPR 2025 · similarity 0.86)
- MeshArt: Generating Articulated Meshes with Structure-Guided Transformers (CVPR 2025 · similarity 0.85)
- A Simple Baseline for Efficient Hand Mesh Reconstruction (CVPR 2024 · 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.