Research map / Video / Action recognition
Online Detection of Action Start in Untrimmed, Streaming Videos
Zheng Shou, Junting Pan, Jonathan Chan, Kazuyuki Miyazawa, Hassan Mansour, Anthony Vetro, Xavier Giro-i-Nieto, Shih-Fu Chang
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 the three proposed training methods (adaptive sampling, temporal consistency, GAN hard negatives) to TSN | ODAS model · ActivityNet (offset 10s) | point-level AS detection mAP (%): 8.18 -> 8.33 | Improved |
| add the three proposed training methods to TSN | ODAS model · ActivityNet (offset 50s) | point-level AS detection mAP (%): 31.39 -> 33.08 | Improved |
| add the three proposed training methods to TSN | ODAS model · ActivityNet (offset 100s) | point-level AS detection mAP (%): 44.15 -> 46.97 | Improved |
| add adaptive sampling, temporal consistency or GAN individually to C3D (ablation) | C3D · THUMOS'14 | point-level AS detection mAP | Improved |
| add adaptive sampling to C3D-TC-GAN (ablation) | C3D-TC-GAN · THUMOS'14 | point-level AS detection mAP | Improved |
| add temporal consistency to C3D-adaptive-GAN (ablation) | C3D-adaptive-GAN · THUMOS'14 | point-level AS detection mAP | Improved |
| add GAN hard negatives to C3D-adaptive-TC (ablation) | C3D-adaptive-TC · THUMOS'14 | point-level AS detection mAP | Improved |
| increase the input window stride to 8 frames (ablation) | Our approach · THUMOS'14 | mAP | No clear effect |
Closest papers by meaning
- StartNet: Online Detection of Action Start in Untrimmed Videos (ICCV 2019 · similarity 0.90)
- Online Action Detection (ECCV 2016 · similarity 0.87)
- MiniROAD: Minimal RNN Framework for Online Action Detection (ICCV 2023 · similarity 0.86)
- Learning to Discriminate Information for Online Action Detection (CVPR 2020 · similarity 0.85)
- Does Video-Text Pretraining Help Open-Vocabulary Online Action Detection? (NeurIPS 2024 · similarity 0.84)
- Context-Enhanced Memory-Refined Transformer for Online Action Detection (CVPR 2025 · similarity 0.83)
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.
