Research map / 3D vision / Human pose estimation
Adversarial Semantic Data Augmentation for Human Pose Estimation
Yanrui Bin, Xuan Cao, Xinya Chen, Yanhao Ge, Ying Tai, Chengjie Wang, Jilin Li, Feiyue Huang, Changxin Gao, Nong Sang
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 Semantic Data Augmentation (random part pasting) (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 90.3 -> 90.8 | Improved |
| add Semantic Data Augmentation on invisible keypoints (ablation) | HRNet-W32 · MPII val (invisible keypoints) | PCKh@0.5 total (%): 74.2 -> 75.4 | Improved |
| add ROR (random occluding and repeating) augmentation (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 90.3 -> 90.5 | Improved |
| add ROR augmentation, knee and ankle (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 87.1 -> 86.9 (knee), 83.3 -> 82.9 (ankle) | Hurt |
| replace ROR with SDA (semantic body part patches) (ablation) | HRNet-W32 + augmentation · MPII val | PCKh@0.5 total (%): 90.5 -> 90.8 | Improved |
| add adversarial generator (ASDA) over random SDA (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 90.8 -> 91.2 | Improved |
| add adversarial generator (ASDA) over SDA, invisible keypoints (ablation) | HRNet-W32 · MPII val (invisible keypoints) | PCKh@0.5 total (%): 75.4 -> 76.1 | Improved |
| add ASDA to baseline, invisible keypoints (ablation) | HRNet-W32 · MPII val (invisible keypoints) | PCKh@0.5 total (%): 74.2 -> 76.1 | Improved |
| increase pasted part number N from 1 to 3 (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 91.2 -> 91.0 | No clear effect |
| increase pasted part number N from 1 to 4 (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 91.2 -> 90.7 | Hurt |
| increase pasted part number N from 1 to 8 (ablation) | HRNet-W32 · MPII val | PCKh@0.5 total (%): 91.2 -> 89.4 | Hurt |
| add ASDA to 8-stacked hourglass | 8-Stacked HG · MPII val | PCKh@0.5 total (%): 90.2 -> 91.1 | Improved |
Closest papers by meaning
- Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation (CVPR 2018 · similarity 0.90)
- Adversarial PoseNet: A Structure-Aware Convolutional Network for Human Pose Estimation (ICCV 2017 · similarity 0.89)
- SemGAN: A Semantic and Hierarchical Adversarial Network for 3D Human Pose Estimation (ECCV 2026 · similarity 0.88)
- A Dual-Augmentor Framework for Domain Generalization in 3D Human Pose Estimation (CVPR 2024 · similarity 0.87)
- When Human Pose Estimation Meets Robustness: Adversarial Algorithms and Benchmarks (CVPR 2021 · similarity 0.87)
- End-To-End Learning of Deformable Mixture of Parts and Deep Convolutional Neural Networks for Human Pose Estimation (CVPR 2016 · similarity 0.86)
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.
