Research map / Graph learning / Knowledge and temporal graphs
Group Activity Prediction with Sequential Relational Anticipation Model
Junwen Chen, Wentao Bao, Yu Kong
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 position regression loss and position prediction (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 84.85 -> 86.85 | Improved |
| add adversarial loss (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 86.30 -> 86.85 | Improved |
| add sequential reconstruction loss (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 86.22 -> 86.85 | Improved |
| use 5 unrolling stages instead of 1 (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 81.47 -> 86.85 | Improved |
| use 5 unrolling stages instead of 2 (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 85.22 -> 86.85 | Improved |
| use 10 unrolling stages instead of 5 (ablation) | SRAM · Volleyball (avg 10/40/70% observed) | accuracy (%): 86.85 -> 86.79 | No clear effect |
| use 5 stages instead of 1 at low observation ratio (ablation) | SRAM · Volleyball (10% observed) | accuracy (%): 70.38 -> 77.86 | Improved |
| use SRAM instead of e-AAPNet with tracklets | SRAM · Collective Activity (50% observed) | accuracy (%): 86.01 -> 92.55 | Improved |
| use SRAM instead of ARG at low observation | SRAM · Volleyball (10% observed) | accuracy (%): 64.82 -> 77.86 | Improved |
| predict positions with SRAM instead of SocialGAN | SRAM · Volleyball position prediction | FDE: 5.32 -> 3.62 | Improved |
| add tracklet information to AAPNet (e-AAPNet) | SRAM · Volleyball (100% observed) | accuracy (%): 80.78 -> 89.01 | Improved |
Closest papers by meaning
- Progressive Relation Learning for Group Activity Recognition (CVPR 2020 · similarity 0.88)
- GroupFormer: Group Activity Recognition With Clustered Spatial-Temporal Transformer (ICCV 2021 · similarity 0.88)
- A Hierarchical Deep Temporal Model for Group Activity Recognition (CVPR 2016 · similarity 0.87)
- Hierarchical Relational Networks for Group Activity Recognition and Retrieval (ECCV 2018 · similarity 0.87)
- Learning Actor Relation Graphs for Group Activity Recognition (CVPR 2019 · similarity 0.87)
- Learning Group Activity Features Through Person Attribute Prediction (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.
