Research map / Video / Action recognition
Else-Net: Elastic Semantic Network for Continual Action Recognition From Skeleton Data
Tianjiao Li, Qiuhong Ke, Hossein Rahmani, Rui En Ho, Henghui Ding, Jun Liu
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 block searching (select best-matched block per elastic unit) (ablation) | Else-Net · NTU RGB+D continual (CS) | ACC: 76.4 -> 84.4 | Improved |
| add block searching (ablation) | Else-Net · NTU RGB+D continual (CS) | FM: 17.5 -> 5.1 | Improved |
| add selective updating via Gumbel Softmax (ablation) | Else-Net · NTU RGB+D continual (CS) | ACC: 77.3 -> 84.4 | Improved |
| add selective updating via Gumbel Softmax (ablation) | Else-Net · NTU RGB+D continual (CS) | FM: 13.4 -> 5.1 | Improved |
| add five body part branches (ablation) | Else-Net · NTU RGB+D continual (CS) | ACC: 83.0 -> 84.4 | Improved |
| add five body part branches (ablation) | Else-Net · NTU RGB+D continual (CS) | FM: 7.4 -> 5.1 | Improved |
| add body part branches (reduce growth rate of learning blocks) (ablation) | Else-Net · NTU RGB+D continual (CS) | growth rate: 1070% -> 68% | Improved |
| increase elastic units from 1 to 3 (ablation) | Else-Net · NTU RGB+D continual (CS) | ACC: 80.2 -> 84.4 | Improved |
| increase elastic units from 3 to 4 (ablation) | Else-Net · NTU RGB+D continual (CS) | ACC: 84.4 -> 84.6 | No clear effect |
| Else-Net versus MS-G3D trained continually | MS-G3D · NTU RGB+D continual (CS) | ACC: 46.3 -> 84.4 | Improved |
| Else-Net versus MS-G3D in offline learning | MS-G3D · NTU RGB+D offline (CS) | ACC: 91.5 -> 91.6 | No clear effect |
Closest papers by meaning
- Adaptive RNN Tree for Large-Scale Human Action Recognition (ICCV 2017 · similarity 0.87)
- InfoGCN: Representation Learning for Human Skeleton-Based Action Recognition (CVPR 2022 · similarity 0.87)
- Recovering Complete Actions for Cross-dataset Skeleton Action Recognition (NeurIPS 2024 · similarity 0.87)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.87)
- Heterogeneous Skeleton-Based Action Representation Learning (CVPR 2025 · similarity 0.86)
- Neuron: Learning Context-Aware Evolving Representations for Zero-Shot Skeleton Action Recognition (CVPR 2025 · similarity 0.86)
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.
