Research map / Video / Action recognition
Hierarchical Nonlinear Orthogonal Adaptive-Subspace Self-Organizing Map Based Feature Extraction for Human Action Recognition
Yang Du, Chunfeng Yuan, Bing Li, Weiming Hu, Hao Yang, Zhikang Fu, Lili Zhao
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 nonlinear orthogonal map (NOASSOM vs ASSOM) (ablation) | hierarchical NOASSOM · HMDB-51 | mean accuracy: 62.9% -> 66.4% | Improved |
| add nonlinear orthogonal map (NOASSOM vs ASSOM) (ablation) | hierarchical NOASSOM · UCF-101 | mean accuracy: 87.4% -> 92.1% | Improved |
| use temporal channel instead of spatial channel (ablation) | NOASSOM · HMDB-51 | mean accuracy: 58.5% -> 63.4% | Improved |
| use temporal channel instead of spatial channel (ablation) | NOASSOM · UCF-101 | mean accuracy: 83.6% -> 88.2% | Improved |
| fuse spatial and temporal channels in third layer (ablation) | NOASSOM · HMDB-51 | mean accuracy: 63.4% -> 66.4% | Improved |
| fuse spatial and temporal channels in third layer (ablation) | NOASSOM · UCF-101 | mean accuracy: 88.2% -> 92.1% | Improved |
| add iDT features to fusion channel input (ablation) | NOASSOM · HMDB-51 | mean accuracy: 66.4% -> 69.3% | Improved |
| add iDT features to fusion channel input (ablation) | NOASSOM · UCF-101 | mean accuracy: 92.1% -> 93.8% | Improved |
| skip unsupervised pre-training on UCF-101 (ablation) | NOASSOM · HMDB-51 | mean accuracy | Hurt |
| NOASSOM vs best hand-crafted iDT+HSV | NOASSOM · HMDB-51 | mean accuracy: 61.1% -> 66.4% | Improved |
| NOASSOM vs TDD+iDT | NOASSOM+iDT · UCF-101 | mean accuracy: 91.5% -> 93.8% | Improved |
| NOASSOM+iDT vs ST-ResNet+iDT | NOASSOM+iDT · UCF-101 | mean accuracy: 94.6% -> 93.8% | Hurt |
Closest papers by meaning
- Hierarchical-aware Orthogonal Disentanglement Framework for Fine-grained Skeleton-based Action Recognition (ICCV 2025 · similarity 0.83)
- A Hierarchical Pose-Based Approach to Complex Action Understanding Using Dictionaries of Actionlets and Motion Poselets (CVPR 2016 · similarity 0.83)
- Action Recognition in Video Using Sparse Coding and Relative Features (CVPR 2016 · similarity 0.82)
- Beyond Covariance: SICE and Kernel Based Visual Feature Representation (IJCV 2020 · similarity 0.82)
- Skeleton-based Action Recognition with Non-linear Dependency Modeling and Hilbert-Schmidt Independence Criterion (AAAI 2025 · similarity 0.81)
- Exploring a Principled Framework for Deep Subspace Clustering (ICLR 2025 · similarity 0.80)
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.
