Research map / Vision-language and multimodal / Human and embodied modeling
Learning Visual Context for Group Activity Recognition
Read the paperCheck 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 |
|---|---|---|---|
| Add TCE to the pose branch (ablation) | HRNet-w32 pose branch on Volleyball · Volleyball dataset | MCA (%): 89.7 -> 92.0 | Helped |
| Add TCE to the global branch (ablation) | Inception-v3 global branch on Volleyball · Volleyball dataset | MCA (%): 89.8 -> 92.0 | Helped |
| Use four rather than three TCE heads in the pose branch (ablation) | HRNet-w32 pose branch on Volleyball · Volleyball dataset | MCA (%): 92.0 -> 91.9 | Hurt |
| Add STBiP to the TCE-equipped pose branch (ablation) | HRNet-w32 pose branch on Volleyball · Volleyball dataset | MCA (%): 92.0 -> 93.3 | Helped |
| Add STBiP to the TCE-equipped global branch (ablation) | Inception-v3 global branch on Volleyball · Volleyball dataset | MCA (%): 91.9 -> 92.9 | Helped |
| Use eight rather than four STBiP heads in the global branch (ablation) | Inception-v3 global branch on Volleyball · Volleyball dataset | MCA (%): 92.9 -> 92.5 | Hurt |
| Fuse branches before TCE instead of fusing branch scores at test time (ablation) | Pose and global branches on Volleyball · Volleyball dataset | MCA (%): 93.8 -> 91.8 | Hurt |
| Fuse branches before STBiP instead of fusing branch scores at test time (ablation) | Pose and global branches on Volleyball · Volleyball dataset | MCA (%): 93.8 -> 92.3 | Hurt |
| Add global visual context and the proposed reasoning modules | VGG16 + HRNet† on Volleyball · Volleyball dataset | MCA (%): 93.5 -> 94.7 | Helped |
| Use the proposed method on VGG16 for collective activity recognition | VGG16 on Collective Activity · Collective Activity dataset | MPCA (%): 93.8 -> 95.4 | Helped |
Closest papers by meaning
- Detector-Free Weakly Supervised Group Activity Recognition (CVPR 2022 · similarity 0.90)
- Actor-Transformers for Group Activity Recognition (CVPR 2020 · similarity 0.88)
- Learning to Evolve Scenes: Reasoning about Human Activities with Scene Graphs (NeurIPS 2026 · similarity 0.85)
- Learning from Synthetic Human Group Activities (CVPR 2024 · similarity 0.85)
- Bi-Causal: Group Activity Recognition via Bidirectional Causality (CVPR 2024 · similarity 0.84)
- Perceptual Group Tokenizer: Building Perception with Iterative Grouping (ICLR 2024 · similarity 0.84)
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.