Research map / Video / Action recognition
Predicting Human Activities Using Stochastic Grammar
Siyuan Qi, Siyuan Huang, Ping Wei, Song-Chun Zhu
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 |
|---|---|---|---|
| use ST-AOG with joint grammar parsing instead of ATCRF for action detection | ST-AOG · CAD-120 detection (action) | micro P/R: 70.3 -> 76.5 | Improved |
| use ST-AOG instead of ATCRF for action detection macro F1 | ST-AOG · CAD-120 detection (action) | macro F1-score: 70.2 -> 76.1 | Improved |
| use ST-AOG instead of ATCRF for affordance detection | ST-AOG · CAD-120 detection (affordance) | micro P/R: 85.4 -> 82.4 | Hurt |
| use ST-AOG instead of ATCRF for affordance detection macro F1 | ST-AOG · CAD-120 detection (affordance) | macro F1-score: 71.9 -> 69.3 | Hurt |
| use ST-AOG instead of ATCRF for action prediction | ST-AOG · CAD-120 prediction (action) | micro P/R: 49.6 -> 55.2 | Improved |
| use ST-AOG instead of ATCRF for action prediction macro F1 | ST-AOG · CAD-120 prediction (action) | macro F1-score: 40.6 -> 56.6 | Improved |
| use ST-AOG instead of ATCRF for affordance prediction | ST-AOG · CAD-120 prediction (affordance) | micro P/R: 67.2 -> 73.5 | Improved |
| use ST-AOG instead of ATCRF for affordance prediction macro F1 | ST-AOG · CAD-120 prediction (affordance) | macro F1-score: 41.4 -> 56.2 | Improved |
| add joint high-level grammar reasoning to low-level detections (one stacking objects video) | ST-AOG · CAD-120 stacking objects video | action detection accuracy: 50.9% -> 86.7% | Improved |
| add joint high-level grammar reasoning to low-level affordance detections (one video) | ST-AOG · CAD-120 stacking objects video | affordance detection accuracy: 84.5% -> 87.3% | Improved |
Closest papers by meaning
- MOMA-LRG: Language-Refined Graphs for Multi-Object Multi-Actor Activity Parsing (NeurIPS 2022 · similarity 0.89)
- Adversarial Generative Grammars for Human Activity Prediction (ECCV 2020 · similarity 0.88)
- A Generalized Earley Parser for Human Activity Parsing and Prediction (TPAMI 2020 · similarity 0.88)
- Anticipating Human Activities Using Object Affordances for Reactive Robotic Response (TPAMI 2016 · similarity 0.87)
- What Actions Are Needed for Understanding Human Actions in Videos? (ICCV 2017 · similarity 0.87)
- Learning to Evolve Scenes: Reasoning about Human Activities with Scene Graphs (NeurIPS 2026 · 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.
