Research map / 3D vision / 3D generation
3D Single-Person Concurrent Activity Detection Using Stacked Relation Network
Yi Wei, Wenbo Li, Yanbo Fan, Linghan Xu, Ming-Ching Chang, Siwei Lyu
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 relation layers in both RN-PP and RN-CLS (vs none) (ablation) | SRN · UCLA concurrent activity | MAP: 0.81 -> 0.93 | Improved |
| Add relation layers in RN-CLS only (vs none) (ablation) | SRN · UCLA concurrent activity | MAP: 0.81 -> 0.89 | Improved |
| Add relation layers in RN-PP (vs none in RN-PP) (ablation) | SRN · UCLA concurrent activity | OAP: 0.85 -> 0.91 | Improved |
| Add T-relation module in RN-CLS (ablation) | SRN · UCLA concurrent activity | MAP: 0.87 -> 0.93 | Improved |
| Add T-relation module to RN-PP as well (ablation) | SRN · UCLA concurrent activity | MAP: 0.93 -> 0.87 | Hurt |
| Add T-relation to RN-PP (ablation) | SRN · UCLA concurrent activity | ER: 0.22 -> 0.24 | Hurt |
| Combine attentive weight with learnable base weight (vs learnable only) (ablation) | SRN · UCLA concurrent activity | MAP: 0.88 -> 0.93 | Improved |
| Combine learnable base weight with attentive weight (vs attentive only) (ablation) | SRN · UCLA concurrent activity | MAP: 0.83 -> 0.93 | Improved |
| Use decompositional class-wise representation instead of unified representation (ablation) | SRN · UCLA concurrent activity | ER: 0.40 -> 0.22 | Improved |
| Use decompositional class-wise representation instead of unified representation (ablation) | SRN · UCLA concurrent activity | MAP: 0.86 -> 0.93 | Improved |
| SRN vs AGCN with same input features | SRN · UCLA concurrent activity | MAP: 0.86 -> 0.93 | Improved |
| SRN vs best baseline on new dataset | SRN · New concurrent activity dataset | OAP: 0.57 -> 0.61 | Improved |
Closest papers by meaning
- Skeleton-based Group Activity Recognition via Spatial-Temporal Panoramic Graph (ECCV 2024 · similarity 0.86)
- Learning by Aligning 2D Skeleton Sequences and Multi-Modality Fusion (ECCV 2024 · similarity 0.86)
- Hierarchical Relational Networks for Group Activity Recognition and Retrieval (ECCV 2018 · similarity 0.86)
- Graph Diffusion Convolutional Network for Skeleton Based Semantic Recognition of Two-Person Actions (TPAMI 2023 · similarity 0.85)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.85)
- BlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition (CVPR 2024 · similarity 0.85)
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.
