Research map / Video / Action recognition
Bottom-Up Temporal Action Localization with Mutual Regularization
Peisen Zhao, Lingxi Xie, Chen Ju, Ya Zhang, Yanfeng Wang, Qi Tian
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 IntraC and InterC consistency losses to the baseline (ablation) | bottom-up TAL baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 42.63 | Improved |
| add IntraC on all three phases (ablation) | baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 41.91 | Improved |
| add IntraC on the continuing phase only (ablation) | baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 40.46 | Improved |
| add IntraC on start and end phases only (ablation) | baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 40.86 | Improved |
| add InterC on both continue-start and continue-end (ablation) | baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 41.23 | Improved |
| add InterC on only one of continue-start or continue-end (ablation) | baseline · THUMOS14 proposals | AR@50 (%): 39.02 -> 40.21 (C&S), 40.64 (C&E) | Improved |
| add InterC on top of IntraC (ablation) | IntraC model · THUMOS14 proposals | AR@50 (%): 41.91 -> 42.63 | Improved |
| add a proposal evaluation network score to start-end probability product (ablation) | proposal ranking · THUMOS14 proposals | AR@50 (%): 42.63 -> 44.23 | Improved |
| increase BaseNet kernel size from 3 to 9 (2 layers) (ablation) | BaseNet · THUMOS14 proposals | AR@50 (%): 39.54 -> 42.63; 11: 42.48, 13: 42.17 | Mixed |
| increase BaseNet depth from 2 to 5 layers (kernel 5) (ablation) | BaseNet · THUMOS14 proposals | AR@50 (%): 40.98 -> 41.56 (3 layers), 41.68 (4), 40.98 (5) | No clear effect |
| add IntraC and InterC to TSA-Net-small (reimplemented) | TSA-Net-small · THUMOS14 proposals | AR@50 (%): 38.32 -> 39.73 | Improved |
| add IntraC and InterC to TSA-Net-large (reimplemented) | TSA-Net-large · THUMOS14 proposals | AR@50 (%): 37.91 -> 39.68 | Improved |
Closest papers by meaning
- Action Sensitivity Learning for Temporal Action Localization (ICCV 2023 · similarity 0.91)
- Temporal Action Localization with Cross Layer Task Decoupling and Refinement (AAAI 2025 · similarity 0.90)
- Test-Time Zero-Shot Temporal Action Localization (CVPR 2024 · similarity 0.90)
- Realigning Confidence with Temporal Saliency Information for Point-Level Weakly-Supervised Temporal Action Localization (CVPR 2024 · similarity 0.89)
- Energy vs. Noise: Towards Robust Temporal Action Localization in Open-World (AAAI 2025 · similarity 0.89)
- Similar Modality Enhancement and Action Consistency Learning for Weakly Supervised Temporal Action Localization (AAAI 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.
