Research map / Video / Action recognition
Relational Action Forecasting
Check 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 recurrent action dynamics modeling with a GRU (ablation) | Independent actor action prediction · AVA, t=1 | mean AP@.5: 7.8 -> 13.1 | Helped |
| Add DR²N learned relational modeling to the GRU (ablation) | Graph-GRU · AVA, t=1 | mean AP@.5: 13.1 -> 14.4 | Helped |
| Replace DR²N relations with equally weighted RN edges (ablation) | Graph-GRU · AVA, t=1 | mean AP@.5: 14.4 -> 12.3 | Hurt |
| Replace DR²N relations with GAT (ablation) | Graph-GRU · AVA, t=1 | mean AP@.5: 14.4 -> 12.3 | Hurt |
| Use jointly trained multi-head classifiers instead of separate single-head models (ablation) | Visual-feature action prediction · AVA, t=1 | mean AP@.5: 7.8 -> 9.4 | Helped |
| Add graph relations using DR²N (ablation) | GRU · J-HMDB 10% | accuracy@K: 52.5 -> 60.6 | Helped |
| Replace DR²N relation modeling with GAT (ablation) | J-HMDB early action prediction · J-HMDB 10% | accuracy@K: 58.1 -> 60.6 | Hurt |
Closest papers by meaning
- Modular Action Concept Grounding in Semantic Video Prediction (CVPR 2022 · similarity 0.89)
- Deep Sequential Context Networks for Action Prediction (CVPR 2017 · similarity 0.88)
- Anticipating Future Relations via Graph Growing for Action Prediction (AAAI 2021 · similarity 0.88)
- End-To-End Learning of Action Detection From Frame Glimpses in Videos (CVPR 2016 · similarity 0.88)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · similarity 0.86)
- Latent Action Pretraining from Videos (ICLR 2025 · 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.