Research map / Video / Action recognition
Modeling Temporal Dynamics and Spatial Configurations of Actions Using Two-Stream Recurrent Neural Networks
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 |
|---|---|---|---|
| Combine temporal and spatial RNN streams | Skeleton action recognition · NTU RGB+D cross subject | accuracy: 66.1 -> 68.6 | Helped |
| Combine temporal and spatial RNN streams | Skeleton action recognition · SBU Interaction | accuracy: 89.0 -> 91.9 | Helped |
| Add 3D transformation augmentation (ablation) | Two-stream RNN · NTU RGB+D cross view | accuracy: 71.7 -> 79.5 | Helped |
| Add 3D transformation augmentation (ablation) | Two-stream RNN · NTU RGB+D cross subject | accuracy: 68.6 -> 71.3 | Helped |
| Use hierarchical rather than stacked temporal RNN (ablation) | Temporal RNN · NTU RGB+D cross subject | accuracy: 66.1 -> 67.8 | Helped |
| Use traversal rather than chain joint ordering (ablation) | Spatial RNN · SBU Interaction | accuracy: 82.2 -> 86.6 | Helped |
| Add a third stacked RNN layer (ablation) | Stacked RNN · NTU RGB+D cross view | accuracy: 68.9 -> 69.2 | Helped |
| Reduce stacked RNN hidden size to 256 per layer (ablation) | Stacked RNN · NTU RGB+D cross view | accuracy: 68.9 -> 68.2 | Hurt |
| Increase stacked RNN hidden size to 1024 per layer (ablation) | Stacked RNN · NTU RGB+D cross view | accuracy: 68.9 -> 68.9 | No clear effect |
| Add a second RNN layer for body parts (ablation) | Hierarchical RNN · NTU RGB+D cross view | accuracy: 70.5 -> 71.4 | Helped |
| Increase body-part hidden size from 128 to 256 (ablation) | Hierarchical RNN · NTU RGB+D cross view | accuracy: 70.5 -> 71.4 | Helped |
| Increase whole-body hidden size from 512 to 1024 (ablation) | Hierarchical RNN · NTU RGB+D cross view | accuracy: 70.5 -> 70.6 | No clear effect |
Closest papers by meaning
- Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition (AAAI 2018 · similarity 0.91)
- Two-Stream Adaptive Graph Convolutional Networks for Skeleton-Based Action Recognition (CVPR 2019 · similarity 0.91)
- A New Representation of Skeleton Sequences for 3D Action Recognition (CVPR 2017 · similarity 0.90)
- An End-to-End Spatio-Temporal Attention Model for Human Action Recognition from Skeleton Data (AAAI 2017 · similarity 0.90)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.89)
- Heterogeneous Skeleton-Based Action Representation Learning (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.