Research map / Video / Action recognition
Weakly-Supervised Action Localization by Generative Attention Modeling
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace the baseline weakly supervised localizer with DGAM | Weakly supervised action localization with I3D features · THUMOS14 test | mAP@0.5: 26.8 -> 28.8 | Helped |
| Replace the baseline weakly supervised localizer with DGAM | Weakly supervised action localization with I3D features · THUMOS14 test | mAP@0.7: 9.0 -> 11.4 | Helped |
| Replace the baseline weakly supervised localizer with DGAM | Weakly supervised action localization · ActivityNet1.2 validation | mAP@AVG: 22.4 -> 24.4 | Helped |
| Add background classification loss (ablation) | Foreground-classification attention model · THUMOS14 | mAP@0.5: 21.5 -> 24.8 | Helped |
| Add self-guided regularization loss (ablation) | Foreground and background classification model · THUMOS14 | mAP@0.5: 24.8 -> 26.7 | Helped |
| Add reconstruction loss and train the conditional VAE (ablation) | DGAM attention model · THUMOS14 | mAP@0.5: 26.7 -> 28.8 | Helped |
| Increase latent dimension from 16 to 128 (ablation) | DGAM conditional VAE · THUMOS14 | mAP@0.5: 26.5 -> 28.8 | Helped |
| Increase latent dimension from 128 to 256 or 512 (ablation) | DGAM conditional VAE · THUMOS14 | mAP@0.5: 28.8 -> 28.3; 28.8 -> 27.7 | Hurt |
| Vary CVAE beta across tested settings (ablation) | DGAM conditional VAE · THUMOS14 | mAP@0.5: 28.2, 28.1, 28.4, 28.8, 28.0, 28.4 | No clear effect |
| Use the new DGAM attention with the old classifier instead of the old attention and classifier (ablation) | THUMOS14 attention/classification module combination · THUMOS14 | mAP@0.5: 26.7 -> 28.8 | Helped |
| Replace the old classifier with the new classifier while retaining the old attention (ablation) | THUMOS14 attention/classification module combination · THUMOS14 | mAP@0.5: 26.7 -> 27.0 | No clear effect |
Closest papers by meaning
- Weakly-supervised Temporal Action Localization by Uncertainty Modeling (AAAI 2021 · similarity 0.92)
- PivoTAL: Prior-Driven Supervision for Weakly-Supervised Temporal Action Localization (CVPR 2023 · similarity 0.92)
- A Hybrid Attention Mechanism for Weakly-Supervised Temporal Action Localization (AAAI 2021 · similarity 0.92)
- Weakly-Supervised Temporal Action Localization by Inferring Salient Snippet-Feature (AAAI 2024 · similarity 0.91)
- Similar Modality Enhancement and Action Consistency Learning for Weakly Supervised Temporal Action Localization (AAAI 2025 · similarity 0.90)
- Bridge the Gap: From Weak to Full Supervision for Temporal Action Localization with PseudoFormer (CVPR 2025 · similarity 0.90)
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.