Research map / Graph learning / Knowledge and temporal graphs
Symbiotic Graph Neural Networks for 3D Skeleton-Based Human Action Recognition and Motion Prediction
Maosen Li, Siheng Chen, Xu Chen, Ya Zhang, Yanfeng Wang, Qi Chuan Tian
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 motion-prediction head with multitasking (ablation) | Sym-GNN · NTU-RGB+D Cross-Subject | accuracy: 89.0% -> 90.1% | Improved |
| add motion-prediction head with multitasking (ablation) | Sym-GNN · NTU-RGB+D Cross-View | accuracy: 95.7% -> 96.4% | Improved |
| add motion-prediction head with multitasking (ablation) | Sym-GNN · Kinetics top-1 | accuracy: 36.4% -> 37.2% | Improved |
| add motion-prediction head with multitasking (ablation) | Sym-GNN · Human 3.6M recognition top-1 | accuracy: 55.2% -> 56.5% | Improved |
| add motion-prediction head with multitasking (ablation) | Sym-GNN · CMU Mocap recognition top-1 | accuracy: 96.6% -> 98.8% | Improved |
| add bone-based dual graph network (ablation) | Sym-GNN · NTU-RGB+D Cross-Subject | accuracy: 87.1% -> 90.1% | Improved |
| add bone-based dual graph network (ablation) | Sym-GNN · CMU Mocap recognition top-1 | accuracy: 93.5% -> 98.8% | Improved |
| add bone-based dual graph network (ablation) | Sym-GNN · Human 3.6M recognition top-1 | accuracy: 53.5% -> 56.5% | Improved |
| add part-scale and actional graphs to joint-scale structural graphs only (ablation) | Sym-GNN (Only J-S) · NTU-RGB+D Cross-Subject | accuracy: 88.3% -> 90.1% | Improved |
| add joint-scale graphs to part-scale graph only (ablation) | Sym-GNN (Only P) · NTU-RGB+D Cross-View | accuracy: 87.3% -> 96.4% | Improved |
| add action-recognition head to motion prediction (ablation) | Sym-GNN · NTU-RGB+D Cross-Subject | PCK@0.05 (%): 92.34 -> 93.48 | Improved |
| add action-recognition head to motion prediction (ablation) | Sym-GNN · Human 3.6M 400 ms average | MAE: 0.94 -> 0.92 | Improved |
Closest papers by meaning
- Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · similarity 0.92)
- Skeleton-Based Action Recognition With Directed Graph Neural Networks (CVPR 2019 · similarity 0.92)
- Learning Graph Convolutional Network for Skeleton-Based Human Action Recognition by Neural Searching (AAAI 2020 · similarity 0.92)
- BlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition (CVPR 2024 · similarity 0.91)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.90)
- Dynamic Semantic-Based Spatial Graph Convolution Network for Skeleton-Based Human Action Recognition (AAAI 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.
