Research map / Detection and segmentation / CNN features
Context-Guided Adaptive Network for Efficient Human Pose Estimation
Lei Zhao, Jun Wen, Pengfei Wang, Nenggan Zheng
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 |
|---|---|---|---|
| Replace ROIPooler with global-context-guided multi-level ROIGuider (ablation) | FPN-50 pose estimator · COCO val2017 | AP: 57.5 -> 60.5 | Improved |
| Use TNet-D1W64 instead of FPN (ablation) | FPN-50 + ROIGuider · COCO val2017 | AP: 60.5 -> 61.9 | Improved |
| Add fixed-ratio box-area extension (ablation) | TNet-D1W64 + ROIGuider · COCO val2017 | AP: 61.9 -> 62.2 | Improved |
| Use adaptive Dichotomy Extended Area instead of the baseline without area extension (ablation) | TNet-D1W64 + ROIGuider · COCO val2017 | AP: 61.9 -> 63.0 | Improved |
| Add Ambiguity Alleviation using center-line heatmaps (ablation) | TNet-D1W64 + ROIGuider + Dichotomy Extended Area · COCO val2017 | AP: 63.0 -> 63.2 | Improved |
| Use cascade ROIGuider (× 2) (ablation) | TNet-D1W64 + ROIGuider + CLB · COCO val2017 | AP: 63.2 -> 64.4 | Improved |
| Train for 55 additional epochs, reaching 150 epochs total (ablation) | Ours (TNet-D1W64) with flip testing · COCO val2017 | AP: 69.5 -> 70.3 | Improved |
Closest papers by meaning
- AdaptivePose: Human Parts as Adaptive Points (AAAI 2022 · similarity 0.87)
- Optimizing Human Pose Estimation Through Focused Human and Joint Regions (AAAI 2025 · similarity 0.87)
- Transformers With Joint Tokens and Local-Global Attention for Efficient Human Pose Estimation (TPAMI 2026 · similarity 0.86)
- Multi-Scale Structure-Aware Network for Human Pose Estimation (ECCV 2018 · similarity 0.86)
- Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimation (ICLR 2023 · similarity 0.86)
- End-to-End Multi-Person Pose Estimation with Pose-Aware Video Transformer (AAAI 2026 · 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.
