Research map / 3D vision / Human pose estimation
DenseRaC: Joint 3D Pose and Shape Estimation by Dense Render-and-Compare
Yuanlu Xu, Song-Chun Zhu, Tony Tung
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 dense render-and-compare losses (ablation) | DenseRaC · H3.6M Protocol #1 | MPJPE: 82.4 -> 79.5 | Improved |
| add synthetic data and own body model after render-and-compare (ablation) | DenseRaC · H3.6M Protocol #1 | MPJPE: 79.5 -> 76.8 | Improved |
| add dense render-and-compare losses (ablation) | DenseRaC · H3.6M Protocol #2 | MPJPE: 53.9 -> 51.4 | Improved |
| add synthetic data and own body model (ablation) | DenseRaC · H3.6M Protocol #2 | MPJPE: 51.4 -> 48.0 | Improved |
| add dense render-and-compare losses (ablation) | DenseRaC · MPI-INF-3DHP Protocol #1 | MPJPE: 123.1 -> 124.9 | Mixed |
| add synthetic data and own body model (ablation) | DenseRaC · MPI-INF-3DHP Protocol #1 | MPJPE: 124.9 -> 114.2 | Improved |
| add adversarial loss (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 133.0 -> 131.5 | Improved |
| add mask loss (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 131.5 -> 122.8 | Improved |
| use dense L_rpj instead of 14-joint-only reprojection (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 122.8 -> 107.9 | Improved |
| add 3D reconstruction loss (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 122.8 -> 86.5 | Improved |
| add regression loss and drop adversarial loss (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 82.4 -> 80.4 | Improved |
| add all loss terms (full model) (ablation) | DenseRaC · MOCA reconstruction | MPJPE: 80.4 -> 80.3 | Improved |
Closest papers by meaning
- DensePose: Dense Human Pose Estimation in the Wild (CVPR 2018 · similarity 0.91)
- SimPose: Effectively Learning DensePose and Surface Normals of People from Simulated Data (ECCV 2020 · similarity 0.90)
- Learning Visibility for Robust Dense Human Body Estimation (ECCV 2022 · similarity 0.89)
- Learning to Estimate 3D Human Pose and Shape From a Single Color Image (CVPR 2018 · similarity 0.89)
- Learning 3D Human Shape and Pose from Dense Body Parts (TPAMI 2020 · similarity 0.89)
- MeshPose: Unifying DensePose and 3D Body Mesh Reconstruction (CVPR 2024 · 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.
