Research map / Label-efficient and robust learning / Recognition and long-tail
InfoGCN: Representation Learning for Human Skeleton-Based Action Recognition
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 |
|---|---|---|---|
| Add the information-bottleneck MMD losses (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 85.1 -> 85.7 | Helped |
| Add context-dependent self-attention topology to learned shared topology (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 84.5 -> 85.1 | Helped |
| Replace static extrinsic topology with inferred context-dependent topology (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 82.8 -> 85.1 | Helped |
| Add marginal-MMD and conditional-marginal-MMD losses (ablation) | Baseline · NTU RGB+D 120 cross-subject | accuracy: +0.3 | Helped |
| Add SA-GC (ablation) | Baseline · NTU RGB+D 120 cross-subject | accuracy: +0.3 | Helped |
| Add MMD losses and SA-GC (ablation) | Baseline · NTU RGB+D 120 cross-subject | accuracy: +0.6 | Helped |
| Add bone representation to joint representation for model ensembling (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 85.1 -> 88.5 | Helped |
| Increase the ensemble from joint + bone to 4 modes (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 88.5 -> 89.4 | Helped |
| Increase the ensemble from 4 modes to 6 modes (ablation) | InfoGCN · NTU RGB+D 120 cross-subject | accuracy: 89.4 -> 89.8 | Helped |
Closest papers by meaning
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.93)
- Graph CNNs with Motif and Variable Temporal Block for Skeleton-Based Action Recognition (AAAI 2019 · similarity 0.92)
- Heterogeneous Skeleton-Based Action Representation Learning (CVPR 2025 · similarity 0.92)
- BlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition (CVPR 2024 · similarity 0.91)
- Dynamic Semantic-Based Spatial Graph Convolution Network for Skeleton-Based Human Action Recognition (AAAI 2024 · similarity 0.90)
- Skeleton-based Action Recognition with Non-linear Dependency Modeling and Hilbert-Schmidt Independence Criterion (AAAI 2025 · 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.