Research map / Video / Action recognition
Skeleton-Based Online Action Prediction Using Scale Selection Network
Jun Liu, Amir Shahroudy, Gang Wang, Ling‐Yu Duan, Alex Chichung Kot
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 multi-level structured skeleton representation instead of coordinate concatenation (ablation) | SSNet · OAD | accuracy at 50%: 79.2 -> 81.3 | Improved |
| add multi-level structured representation (ablation) | SSNet · ChaLearn Gesture | accuracy at 90%: 65.9 -> 69.1 | Improved |
| add multi-level structured representation (ablation) | SSNet · PKUMMD | accuracy at 50%: 68.5 -> 74.1 | Improved |
| add multi-level structured representation (ablation) | SSNet · G3D | accuracy at 90%: 82.0 -> 83.7 | Improved |
| add multi-level structured representation (ablation) | SSNet · OAD | SL-Score: 0.69 -> 0.71 | Improved |
| use learned scale selection instead of fixed scale windows (FSNet-MultiNet) (ablation) | SSNet · OAD | accuracy at 90%: 81.6 -> 82.8 | Improved |
| use regressed scales instead of ground-truth scales (ablation) | SSNet · OAD | accuracy at 10%: 66.7 -> 65.8 | No clear effect |
| use fixed window of 255 instead of 63 in FSNet (ablation) | FSNet · OAD | accuracy at 90%: 78.9 -> 71.2 | Hurt |
| increase conv layers from 8 to 14 (ablation) | SSNet · OAD | prediction accuracy: 75.2 -> 80.6 | Improved |
| increase conv layers from 14 to 16 (ablation) | SSNet · OAD | prediction accuracy: 80.6 -> 80.6 | No clear effect |
| add skip connections to class and distance heads (ablation) | SSNet · OAD | accuracy: about 1.5% | Improved |
| add batch normalization (ablation) | SSNet · OAD | accuracy | No clear effect |
Closest papers by meaning
- SSNet: Scale Selection Network for Online 3D Action Prediction (CVPR 2018 · similarity 0.96)
- Online Human Action Detection Using Joint Classification-Regression Recurrent Neural Networks (ECCV 2016 · similarity 0.91)
- A New Representation of Skeleton Sequences for 3D Action Recognition (CVPR 2017 · similarity 0.89)
- Modeling Temporal Dynamics and Spatial Configurations of Actions Using Two-Stream Recurrent Neural Networks (CVPR 2017 · similarity 0.89)
- Graph CNNs with Motif and Variable Temporal Block for Skeleton-Based Action Recognition (AAAI 2019 · similarity 0.88)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.88)
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.
