Research map / 3D vision / Human pose estimation
3D Human Shape and Pose from a Single Low-Resolution Image with Self-Supervised Learning
Xiangyu Xu, Hao Chen, Francesc Moreno-Noguer, László A. Jeni, Fernando De la Torre
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 |
|---|---|---|---|
| use resolution-aware network instead of plain ResNet50 (ablation) | baseline with basic loss · 3DPW (res 32) | MPJPE: 143.63 -> 135.29 | Improved |
| add self-supervision loss (ablation) | resolution-aware network · 3DPW (res 32) | MPJPE: 135.29 -> 124.23 | Improved |
| add self-supervision loss (ablation) | baseline network · 3DPW (res 32) | MPJPE: 143.63 -> 128.88 | Improved |
| add resolution-aware network on top of self-supervision (ablation) | baseline + SS · 3DPW (res 32) | MPJPE: 128.88 -> 124.23 | Improved |
| add contrastive feature loss (ablation) | RA+SS · 3DPW (res 32) | MPJPE: 124.23 -> 117.12 | Improved |
| add MSE feature loss (ablation) | RA+SS · 3DPW (res 32) | MPJPE: 124.23 -> 124.85 | No clear effect |
| add cosine distance feature loss (ablation) | RA+SS · 3DPW (res 32) | MPJPE: 124.23 -> 125.08 | No clear effect |
| add progressive training (ablation) | full model · 3DPW (res 32) | MPJPE: 127.05 -> 117.12 | Improved |
| use directional self-supervision instead of original (ablation) | full model · 3DPW (res 32) | MPJPE: 156.25 -> 117.12 | Improved |
| add hierarchical supervision from all resolutions (ablation) | full model · 3DPW (res 32) | MPJPE: 122.01 -> 117.12 | Improved |
| use proposed method instead of FeaEN baseline | RSC-Net · 3DPW (res 32) | MPJPE: 143.51 -> 117.12 | Improved |
| use proposed method instead of SPIN | RSC-Net · MPI-INF-3DHP (res 32) | MPJPE: 127.27 -> 115.80 | Improved |
Closest papers by meaning
- 3D Human Pose, Shape and Texture from Low-Resolution Images and Videos (TPAMI 2021 · similarity 0.98)
- Learning to Estimate 3D Human Pose and Shape From a Single Color Image (CVPR 2018 · similarity 0.89)
- 3D Human Pose Machines with Self-supervised Learning (TPAMI 2019 · similarity 0.89)
- Convolutional Mesh Regression for Single-Image Human Shape Reconstruction (CVPR 2019 · similarity 0.88)
- Self-Supervised 3D Human Pose Estimation via Part Guided Novel Image Synthesis (CVPR 2020 · similarity 0.88)
- Coarse-To-Fine Volumetric Prediction for Single-Image 3D Human Pose (CVPR 2017 · similarity 0.88)
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.
