Research map / Video / Action recognition
Decompose and Transfer: CoT-Prompting Enhanced Alignment for Open-Vocabulary Temporal Action Detection
Sa Zhu, Wanqian Zhang, Lin Wang, Xiaohua Chen, Chenxu Cui, Jinchao Zhang, Bo Li
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 CoT phase decomposition (CSD) (ablation) | baseline · THUMOS14 (50% seen / 50% unseen) | avg mAP: 40.3 -> 42.1 | Improved |
| add text-infused foreground filtering (TIF) on top of CSD (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 42.1 -> 43.6 | Improved |
| add adaptive phase-wise alignment (APA) (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 43.6 -> 46.9 | Improved |
| add all three modules (CSD, TIF, APA) in the 75/25 split (ablation) | baseline · THUMOS14 (75% seen / 25% unseen) | avg mAP: 45.9 -> 52.1 | Improved |
| increase phases from one to four (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 42.5 -> 46.9 | Improved |
| increase phases from four to six (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 46.9 -> 47.8 | No clear effect |
| add text-infused filtering instead of none (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 44.3 -> 46.9 | Improved |
| use text-infused instead of static filtering (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 45.5 -> 46.9 | Improved |
| use adaptive weighting instead of simple averaging of phase alignments (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 45.8 -> 46.9 | Improved |
| phase-wise alignment instead of global merged-description alignment (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 44.0 -> 46.9 | Improved |
| swap LLM from GPT-4o to Qwen3 (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 46.9 -> 46.2 | No clear effect |
| swap LLM from GPT-4o to GPT-4 (ablation) | PDA · THUMOS14 (50% seen / 50% unseen) | avg mAP: 46.9 -> 46.6 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DyFADet — DyFADet: Dynamic Feature Aggregation for Temporal Action Detection (ECCV 2024) · better on 2 of 2 shared results
- TriDet — TriDet: Temporal Action Detection With Relative Boundary Modeling (CVPR 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Does Video-Text Pretraining Help Open-Vocabulary Online Action Detection? (NeurIPS 2024 · similarity 0.88)
- Kronecker Mask and Interpretive Prompts are Language-Action Video Learners (ICLR 2025 · similarity 0.88)
- Align Before Adapt: Leveraging Entity-to-Region Alignments for Generalizable Video Action Recognition (CVPR 2024 · similarity 0.88)
- Dual DETRs for Multi-Label Temporal Action Detection (CVPR 2024 · similarity 0.88)
- Weakly Supervised Temporal Action Localization via Dual-Prior Collaborative Learning Guided by Multimodal Large Language Models (CVPR 2025 · similarity 0.88)
- Storyboard-guided Alignment for Fine-grained Video Action Recognition (NeurIPS 2025 · similarity 0.87)
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.
