Research map / Video / Action recognition
MOMA: Multi-Object Multi-Actor Activity Parsing
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 |
|---|---|---|---|
| use hypergraph (HGAP) instead of pairwise GCN for graph branch | HGAP · MOMA activity classification | mAP %: 70 -> 73.9 | Helped |
| use hypergraph instead of pairwise GCN | HGAP · MOMA sub-activity classification | mAP %: 37.4 -> 42.5 | Helped |
| use hypergraph instead of pairwise GCN | HGAP · MOMA atomic action classification | mAP %: 29.4 -> 29.2 | No clear effect |
| use hypergraph instead of pairwise GCN | HGAP · MOMA atomic action localization | mAP %: 29.1 -> 30.3 | Helped |
| use hypergraph instead of pairwise GCN | HGAP · MOMA role classification | mAP %: 64.7 -> 46.4 | Hurt |
| add X3D video stream plus hypergraph vs X3D + GCN | HGAP · MOMA activity classification | mAP %: 71.8 -> 73.9 | Helped |
| add hypergraph branch to X3D | X3D · MOMA activity classification | mAP %: 70.9 -> 73.9 | Helped |
| add hypergraph branch to X3D | X3D · MOMA atomic action classification | mAP %: 27.1 -> 29.2 | Helped |
| use ground-truth hypergraph structure (Oracle) (ablation) | HGAP · MOMA activity classification | mAP %: 73.9 -> 88.4 | Helped |
| use ground-truth hypergraph structure (Oracle) (ablation) | HGAP · MOMA role classification | mAP %: 46.4 -> 54.3 | Helped |
| use HGAP instead of ResNet-50 without graph | HGAP · MOMA role classification | mAP %: 37.9 -> 46.4 | Helped |
Closest papers by meaning
- MOMA-LRG: Language-Refined Graphs for Multi-Object Multi-Actor Activity Parsing (NeurIPS 2022 · similarity 0.92)
- Action Genome: Actions As Compositions of Spatio-Temporal Scene Graphs (CVPR 2020 · similarity 0.89)
- Learning Actor Relation Graphs for Group Activity Recognition (CVPR 2019 · similarity 0.88)
- Understanding Multi-Task Activities from Single-Task Videos (CVPR 2025 · similarity 0.86)
- Open-Vocabulary Video Relation Extraction (AAAI 2024 · similarity 0.85)
- Learning to Evolve Scenes: Reasoning about Human Activities with Scene Graphs (NeurIPS 2026 · 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.