Research map / Video / Action recognition
Learning Dynamics as Feedback: An Adaptive Entropy Flow Dynamics Framework for Long-tailed Human Action Recognition
Yuan Dong, Zhe Zhao, Liheng Yu, Di Wu, Pengkun 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 E-Adapt (ablation) | CE baseline CTR-GCN · NTU-60-LT | overall accuracy: 74.4 -> 77.9 | Improved |
| Add CS-Mix (ablation) | CE baseline CTR-GCN · NTU-60-LT | overall accuracy: 74.4 -> 78.6 | Improved |
| Combine E-Adapt and CS-Mix (ablation) | CE baseline · NTU-60-LT | Few-shot accuracy: 63.8 -> 78.5 | Improved |
| Add E-Adapt (ablation) | CE baseline · NTU-60-LT | Many-shot accuracy: 86.4 -> 85.1 | Hurt |
| Drive adaptation with entropy flow instead of absolute entropy (ablation) | E-Adapt with Mixup · NTU-60-LT | overall accuracy: 76.1 -> 77.9 | Improved |
| Use CS-Mix instead of Mixup (ablation) | E-Adapt · NTU-60-LT | overall accuracy: 77.9 -> 81.5 | Improved |
| Use CS-Mix instead of ShapMix (ablation) | E-Adapt · NTU-60-LT | overall accuracy: 81.0 -> 81.5 | Improved |
| Vary eta and gamma (ablation) | AEED · NTU-60-LT | overall accuracy: 80.4 to 81.5 | No clear effect |
| Add AEED to balanced training | CTR-GCN · NTU-60 X-Sub balanced | top-1 accuracy: 92.4 -> 93.8 | No clear effect |
| Add AEED controller to baselines | ResNet-32 · CIFAR-100-LT | top-1 accuracy: 38.8 -> 42.5 | Improved |
| Add AEED controller to BERT-base | BERT-base · Amazon-531 | sample-F1: 42.1 -> 44.9 | Improved |
| AEED vs Shap-Mix on Many-shot | Shap-Mix · NTU-60-LT | Many-shot accuracy: 86.8 -> 85.5 | Hurt |
Closest papers by meaning
- InfoGCN++: Learning Representation by Predicting the Future for Online Skeleton-Based Action Recognition (TPAMI 2024 · similarity 0.85)
- VA-AR: Learning Velocity-Aware Action Representations with Mixture of Window Attention (AAAI 2025 · similarity 0.85)
- Efficient Action Recognition via Dynamic Knowledge Propagation (ICCV 2021 · similarity 0.85)
- Bayesian Hierarchical Dynamic Model for Human Action Recognition (CVPR 2019 · similarity 0.85)
- Learning Causal Domain-Invariant Temporal Dynamics for Few-Shot Action Recognition (ICML 2024 · similarity 0.84)
- SeFAR: Semi-supervised Fine-grained Action Recognition with Temporal Perturbation and Learning Stabilization (AAAI 2025 · similarity 0.84)
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.
