Research map / Video / Action recognition
Online Real-Time Multiple Spatiotemporal Action Localisation and Prediction
Gurkirt Singh, Suman Saha, Michael Sapienza, Philip H. S. Torr, Fabio Cuzzolin
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 real-time flow stream to appearance stream (union-set) (ablation) | appearance-only SSD detector · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 40.9 -> 43.0 | Improved |
| use accurate flow instead of real-time flow in fusion (union-set) (ablation) | A+RTF · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 43.0 -> 46.3 | Improved |
| add real-time flow stream to appearance stream (boost-fusion) (ablation) | appearance-only SSD detector · J-HMDB-21 (3 splits, video mAP) | mAP @ IoU 0.5: 59.7 -> 65.0 | Improved |
| use accurate flow instead of real-time flow in fusion (boost-fusion) (ablation) | A+RTF · J-HMDB-21 (3 splits, video mAP) | mAP @ IoU 0.5: 65.0 -> 72.0 | Improved |
| use union-set fusion instead of boost-fusion (ablation) | A+RTF · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 41.9 -> 43.0 | Improved |
| use union-set fusion instead of boost-fusion (ablation) | A+AF · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 44.0 -> 46.3 | Improved |
| use boost-fusion instead of union-set fusion (ablation) | A+RTF · J-HMDB-21 (3 splits, video mAP) | mAP @ IoU 0.5: 63.9 -> 65.0 | Improved |
| use boost-fusion instead of union-set fusion (ablation) | A+AF · J-HMDB-21 (3 splits, video mAP) | mAP @ IoU 0.5: 70.1 -> 72.0 | Improved |
| use online greedy tube generation instead of offline tube generation (ablation) | SSD detections + A+AF · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 43.3 -> 46.3 | Improved |
| use online greedy tube generation instead of offline tube generation (ablation) | SSD detections + A+AF · J-HMDB-21 (3 splits, video mAP) | mAP @ IoU 0.5: 71.1 -> 72.0 | Improved |
| use online greedy tube generation instead of offline tube generation (ablation) | SSD detections + A+AF · UCF101-24 split1 (video mAP) | mAP @ IoU 0.2: 71.7 -> 73.5 | Mixed |
| replace accurate flow with real-time flow in fusion (ablation) | A+AF (union-set) · UCF101-24 split1 (video mAP) | mAP @ IoU 0.5: 46.3 -> 43.0 | Hurt |
Closest papers by meaning
- CFAD: Coarse-to-Fine Action Detector for Spatiotemporal Action Localization (ECCV 2020 · similarity 0.89)
- Predicting the Where and What of Actors and Actions Through Online Action Localization (CVPR 2016 · similarity 0.88)
- A Sliding Window Scheme for Online Temporal Action Localization (ECCV 2022 · similarity 0.88)
- End-to-End Spatio-Temporal Action Localisation with Video Transformers (CVPR 2024 · similarity 0.88)
- Action Tubelet Detector for Spatio-Temporal Action Localization (ICCV 2017 · similarity 0.88)
- Improving Action Localization by Progressive Cross-Stream Cooperation (CVPR 2019 · similarity 0.88)
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.
