Research map / 3D vision / Human pose estimation
Convolutional Pose Machines
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 |
|---|---|---|---|
| Replace pose-machine modules with convolutional networks | Pose Machine · LSP | PCK@0.1: +42.4 percentage points | Helped |
| Replace pose-machine modules with convolutional networks | Pose Machine · LSP | PCK@0.2: +30.9 percentage points | Helped |
| Add long-range receptive field on belief maps (ablation) | Convolutional Pose Machine · FLIC | accuracy | Helped |
| Train jointly from scratch with intermediate supervision (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | Helped |
| Remove intermediate supervision (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | Hurt |
| Train stages stage-wise, then stack them (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | Hurt |
| Initialize joint training from stage-wise weights (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | Helped |
| Increase the number of stages up to five (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | Helped |
| Add a sixth stage (ablation) | Convolutional Pose Machine · LSP person-centric | PCK | No clear effect |
| Add a Gaussian center heat-map for the primary subject to subsequent stages (ablation) | Convolutional Pose Machine · MPII Human Pose | PCKh-0.5 | Helped |
| Add LSP training data (ablation) | CPM · MPII Human Pose | PCKh-0.5: 87.95% -> 88.52% | Helped |
| Add MPII training data (ablation) | CPM · LSP test | PCK: 84.32% -> 90.5% | Helped |
Closest papers by meaning
- Pose Prior Learner: Unsupervised Categorical Prior Learning for Pose Estimation (ICLR 2026 · similarity 0.90)
- A Comparative Analysis and Study of Multiview CNN Models for Joint Object Categorization and Pose Estimation (ICML 2016 · similarity 0.90)
- DSC-PoseNet: Learning 6DoF Object Pose Estimation via Dual-Scale Consistency (CVPR 2021 · similarity 0.89)
- RotationNet: Joint Object Categorization and Pose Estimation Using Multiviews From Unsupervised Viewpoints (CVPR 2018 · similarity 0.89)
- Learning Temporal Pose Estimation from Sparsely-Labeled Videos (NeurIPS 2019 · similarity 0.89)
- LSTM Pose Machines (CVPR 2018 · 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.