Research map / Graph learning / Knowledge and temporal graphs
Part-Level Graph Convolutional Network for Skeleton-Based Action Recognition
Linjiang Huang, Yan Huang, Wanli Ouyang, Liang Wang
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 PR block (ablation) | ST-GCN with concatenated input · NTU RGB+D CS | accuracy: 84.0 -> 88.1 | Improved |
| add PR block instead of a second ST-GCN (same params) (ablation) | ST-GCN · NTU RGB+D CS | accuracy: 86.5 -> 88.1 | Improved |
| keep joint-level ST-GCN alongside PR block (ablation) | PR block · NTU RGB+D CS | accuracy: 84.2 -> 88.1 | Improved |
| use pooling matrix P^T P instead of free learnable matrix (ablation) | PR block · NTU RGB+D CS | accuracy: 84.9 -> 88.1 | Improved |
| add orthogonal constraint on pooling matrix (ablation) | PR block · NTU RGB+D CS | accuracy: 85.0 -> 88.1 | Improved |
| learned instead of fixed hand-defined body parts (ablation) | PR block · NTU RGB+D CS | accuracy: 87.1 -> 88.1 | Improved |
| add normalization matrix (ablation) | PA block · NTU RGB+D CV | accuracy: 92.0 -> 93.5 | Improved |
| attention-based instead of orthogonal unpooling (ablation) | PA block · NTU RGB+D CV | accuracy: 91.0 -> 93.5 | Improved |
| shared pooling across frames (temporal alignment) (ablation) | hierarchical pooling net · NTU RGB+D CS | accuracy: 84.7 -> 86.7 | Improved |
| full PL-GCN vs ST-GCN | ST-GCN · NTU RGB+D | accuracy: +7.7% CS, +6.7% CV | Improved |
Closest papers by meaning
- Actional-Structural Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · similarity 0.93)
- Learning Graph Convolutional Network for Skeleton-Based Human Action Recognition by Neural Searching (AAAI 2020 · similarity 0.93)
- Skeleton-Based Action Recognition With Shift Graph Convolutional Network (CVPR 2020 · similarity 0.92)
- Dynamic Semantic-Based Spatial Graph Convolution Network for Skeleton-Based Human Action Recognition (AAAI 2024 · similarity 0.92)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.90)
- BlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition (CVPR 2024 · similarity 0.90)
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.
