Research map / Adversarial robustness and security / Attacks on text and graphs
When Human Pose Estimation Meets Robustness: Adversarial Algorithms and Benchmarks
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 |
|---|---|---|---|
| Apply AdvMix adversarial augmentation and knowledge distillation | HRNet-W32 · COCO-C | mPC: 53.0 -> 55.5 | Helped |
| Apply AdvMix adversarial augmentation and knowledge distillation | HRNet-W32 · COCO | AP*: 74.4 -> 74.7 | No clear effect |
| Apply AdvMix adversarial augmentation and knowledge distillation | HrHRNet-W32 · COCO-C | mPC: 39.9 -> 45.4 | Helped |
| Apply AdvMix adversarial augmentation and knowledge distillation | HRNet-W32 · MPII-C | mPC: 80.1 -> 83.9 | Helped |
| Apply AdvMix adversarial augmentation and knowledge distillation | HRNet-W32 · MPII | PCKh*: 90.3 -> 90.5 | No clear effect |
| Use per-pixel learned mixing rather than sequential, equal, or Dirichlet mixing (ablation) | HRNet-W32 with augmentation composition · COCO-C | mPC: 54.6; 54.2; 54.4; 55.1 -> 55.9 | Helped |
| Generate per-pixel instead of per-image mixing weights (ablation) | AdvMix on HRNet-W32 · COCO-C | mPC: 55.1 -> 55.9 | Helped |
| Add knowledge distillation (ablation) | AdvMix with HRNet-W32 · MPII-C | mPC: 81.7 -> 83.9 | Helped |
| Add knowledge distillation (ablation) | AdvMix with HRNet-W32 · COCO-C | mPC: 55.9 -> 55.9 | No clear effect |
| Train with stylized images in addition to AdvMix | HRNet-W32 · COCO-C | mPC: 55.9 -> 56.5 | Helped |
| Augment training with selected test-time corruption types | HRNet-W32 · COCO-C | AP*: 76.6 -> 75.4 | Hurt |
| Use a larger-capacity backbone | Pose estimators evaluated on COCO-C · COCO-C | robustness | Helped |
Closest papers by meaning
- PoseExaminer: Automated Testing of Out-of-Distribution Robustness in Human Pose and Shape Estimation (CVPR 2023 · similarity 0.88)
- A Dual-Augmentor Framework for Domain Generalization in 3D Human Pose Estimation (CVPR 2024 · similarity 0.86)
- PoseIRM: Enhance 3D Human Pose Estimation on Unseen Camera Settings via Invariant Risk Minimization (CVPR 2024 · similarity 0.85)
- On the Calibration of Human Pose Estimation (ICML 2024 · similarity 0.85)
- Towards Robust 3D Pose Transfer with Adversarial Learning (CVPR 2024 · similarity 0.84)
- TokenHMR: Advancing Human Mesh Recovery with a Tokenized Pose Representation (CVPR 2024 · similarity 0.84)
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.