Research map / 3D vision / Human pose estimation
Reconstructing 3D Human Pose by Watching Humans in the Mirror
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 mirror symmetry constraint L_s (ablation) | Ours without L_s and L_n · Mirror pose reconstruction evaluation | MPJPE: 53.81 -> 39.52 | Helped |
| Add mirror symmetry constraint L_s (ablation) | Ours without L_s and L_n · Mirror pose reconstruction evaluation | PA-MPJPE: 34.48 -> 33.24 | Helped |
| Add mirror symmetry constraint L_s (ablation) | Ours without L_s and L_n · Mirror pose reconstruction evaluation | MRPE: 108.43 -> 101.91 | Helped |
| Add mirror normal constraint L_n (ablation) | Ours without L_n · Mirror pose reconstruction evaluation | MPJPE: 39.52 -> 38.77 | Helped |
| Add mirror normal constraint L_n (ablation) | Ours without L_n · Mirror pose reconstruction evaluation | PA-MPJPE: 33.24 -> 32.96 | Helped |
| Add mirror normal constraint L_n (ablation) | Ours without L_n · Mirror pose reconstruction evaluation | MRPE: 101.91 -> 93.22 | Helped |
| Train MeshNet with Mirrored-Human instead of the SMPLify-X-labeled Mirrored-Human baseline | MeshNet · 3DPW test | MPJPE: 90.0 -> 85.1 | Helped |
| Train MeshNet with Mirrored-Human instead of the SMPLify-X-labeled Mirrored-Human baseline | MeshNet · 3DPW test | PA-MPJPE: 57.5 -> 54.8 | Helped |
| Train MeshNet with Mirrored-Human annotations | MeshNet · Human3.6M | MPJPE: 54.7 -> 53.6 | Helped |
| Train MeshNet with Mirrored-Human annotations | MeshNet · Human3.6M | PA-MPJPE: 41.7 -> 41.0 | Helped |
| Train Moon with Mirrored-Human | Moon · MuPoTS-3D | AP_root^25: 31.0 -> 42.2 | Helped |
| Train Moon with Mirrored-Human | Moon · MuPoTS-3D | PCK_rel: 81.8 -> 82.3 | Helped |
Closest papers by meaning
- Coherent Reconstruction of Multiple Humans From a Single Image (CVPR 2020 · similarity 0.87)
- Reconstructing Humans with a Biomechanically Accurate Skeleton (CVPR 2025 · similarity 0.86)
- BLADE: Single-view Body Mesh Estimation through Accurate Depth Estimation (CVPR 2025 · similarity 0.86)
- Learning Basis Representation to Refine 3D Human Pose Estimations (AAAI 2019 · similarity 0.86)
- Monocular 3D Pose and Shape Estimation of Multiple People in Natural Scenes - The Importance of Multiple Scene Constraints (CVPR 2018 · similarity 0.86)
- Mirror Surface Reconstruction Under an Uncalibrated Camera (CVPR 2016 · 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.