Research map / Video / Action recognition
Single Image Action Recognition Using Semantic Body Part Actions
Zhichen Zhao, Huimin Ma, Shaodi You
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 semantic part features (ablation) | ResNet-50 body classifier · PASCAL VOC 2012 Action (val) | mAP: 81.4 -> 83.2 | Improved |
| add semantic part features (ablation) | ResNet-50 body classifier · Stanford-40 | mAP: 80.2 -> 82.4 | Improved |
| add LDA-based part selection (top 2 parts) (ablation) | part+Nb+Np · PASCAL VOC 2012 Action (val) | mAP: 83.2 -> 84.0 | Improved |
| add LDA-based part selection (top 2 parts) (ablation) | part+Nb+Np · Stanford-40 | mAP: 82.4 -> 83.4 | Improved |
| add part action network Np features to body network Nb features (ablation) | part features · PASCAL VOC 2012 Action (val) | mAP: 82.0 -> 83.2 | Improved |
| add part action network Np features to body network Nb features (ablation) | part features · Stanford-40 | mAP: 81.2 -> 82.4 | Improved |
| add body network Nb to part-action-only features (ablation) | part features · PASCAL VOC 2012 Action (val) | mAP: 81.8 -> 83.2 | Improved |
| use Nb instead of Np for part features (ablation) | part features · PASCAL VOC 2012 Action (val) | mAP: 81.8 -> 82.0 | Improved |
| use part-action network trained on Stanford-40 only (ablation) | Np · PASCAL VOC 2012 Action (val) | mAP: 81.4 -> 81.8 | Improved |
| use semantic body part actions instead of ResNet-50 alone | Ours · PASCAL VOC 2012 Action (test) | mAP: 82.7 -> 86.4 | Improved |
| use semantic body part actions instead of Action Mask | Ours · Stanford-40 | mAP: 80.8 -> 83.4 | Improved |
| use semantic body part actions instead of ResNet-50 on walking | Ours · PASCAL VOC 2012 Action (val) | AP (walking): 53.9 -> 61.9 | Improved |
Closest papers by meaning
- Part-Level Graph Convolutional Network for Skeleton-Based Action Recognition (AAAI 2020 · similarity 0.87)
- Semantics-Guided Neural Networks for Efficient Skeleton-Based Human Action Recognition (CVPR 2020 · similarity 0.87)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.87)
- UCF-STAR: A Large Scale Still Image Dataset for Understanding Human Actions (AAAI 2020 · similarity 0.86)
- Else-Net: Elastic Semantic Network for Continual Action Recognition From Skeleton Data (ICCV 2021 · similarity 0.86)
- Symbiotic Graph Neural Networks for 3D Skeleton-Based Human Action Recognition and Motion Prediction (TPAMI 2021 · 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.
