Research map / 3D vision / Human pose estimation
Graph-PCNN: Two Stage Human Pose Estimation with Graph Pose Refinement
Jian Wang, Xiang Long, Yuan Gao, Errui Ding, Shilei Wen
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 two-stage framework (stage2 regression) over simple baseline | Graph-PCNN (R50) · COCO val2017 128x96 | AP: 59.3 -> 64.6 | Improved |
| add two-stage framework (stage2 regression) over simple baseline | Graph-PCNN (R50) · COCO val2017 256x192 | AP: 70.4 -> 72.6 | Improved |
| add two-stage framework (stage2 regression) over simple baseline | Graph-PCNN (R50) · COCO val2017 384x288 | AP: 72.2 -> 73.6 | Improved |
| train with Graph-PCNN and use only stage1 heatmap output | Graph-PCNN stage1 (R50) · COCO val2017 128x96 | stage1 AP: 59.3 -> 61.1 | Improved |
| train with Graph-PCNN and use only stage1 heatmap output | Graph-PCNN stage1 (R50) · COCO val2017 384x288 | stage1 AP: 72.2 -> 72.7 | Improved |
| add graph pose refinement instead of struct-agnostic module (ablation) | Graph-PCNN (R50) · COCO val2017 128x96 | stage2 AP: 63.8 -> 64.6 | Improved |
| add graph pose refinement instead of struct-agnostic module (ablation) | Graph-PCNN (R50) · COCO val2017 128x96 | stage1 AP: 60.7 -> 61.1 | Improved |
| use reliability-weighted messages (with heat response) instead of all weights set to 1 (ablation) | GPR · COCO val2017 128x96 | stage2 AP: 52.1 -> 64.6 | Improved |
| add heat response factor to neighbour weights (ablation) | GPR · COCO val2017 128x96 | stage2 AP: 64.5 -> 64.6 | Improved |
| add guided points shuffling during training (ablation) | GPR · COCO val2017 128x96 | stage2 AP: 64.3 -> 64.6 | Improved |
| replace DARK decoding with Graph-PCNN regression refinement | HRNet-W32 128x96 · COCO val2017 | AP: 70.7 -> 71.5 | Improved |
| replace DARK decoding with Graph-PCNN regression refinement | ResNet-152 128x96 · COCO val2017 | AP: 63.1 -> 66.1 | Improved |
Later papers that compared with this method
1 later accepted paper reports Graph-PCNN as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Self-Constrained Inference Optimization on Structural Groups for Human Pose Estimation (ECCV 2022 · better on 2 of 2 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PoseFix — PoseFix: Model-Agnostic General Human Pose Refinement Network (CVPR 2019) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Heatmap Distribution Matching for Human Pose Estimation (NeurIPS 2022 · similarity 0.88)
- Neural Localizer Fields for Continuous 3D Human Pose and Shape Estimation (NeurIPS 2024 · similarity 0.87)
- Graph-Based 3D Multi-Person Pose Estimation Using Multi-View Images (ICCV 2021 · similarity 0.87)
- ProbPose: A Probabilistic Approach to 2D Human Pose Estimation (CVPR 2025 · similarity 0.86)
- Continuous Heatmap Regression for Pose Estimation via Implicit Neural Representation (NeurIPS 2024 · similarity 0.86)
- On the Calibration of Human Pose Estimation (ICML 2024 · similarity 0.85)
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.
