Research map / Video / Action recognition
Dual-Evidential Learning for Weakly-Supervised Temporal Action Localization
Mengyuan Chen, Junyu Gao, Shicai Yang, Changsheng Xu
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 traditional EDL loss (ablation) | CO2-Net · THUMOS14 | mAP@Avg (%): 44.5 -> 44.9 | Improved |
| add balanced evidence label weighting (Eq. 8) (ablation) | CO2-Net + traditional EDL · THUMOS14 | mAP@Avg (%): 44.9 -> 45.2 | Improved |
| add video-level uncertainty weighting (generalized EDL) (ablation) | CO2-Net + balanced EDL · THUMOS14 | mAP@Avg (%): 45.2 -> 45.6 | Improved |
| add snippet-uncertainty-guided progressive complementary loss (ablation) | CO2-Net + generalized EDL · THUMOS14 | mAP@Avg (%): 45.6 -> 46.4 | Improved |
| add traditional EDL loss (ablation) | CO2-Net · THUMOS14 | mAP@IoU0.1 / 0.5 (%): 68.9 vs 69.7 at IoU 0.1; 39.0 vs 38.2 at IoU 0.5 | Mixed |
| use full DELU instead of CO2-Net | CO2-Net · ActivityNet1.2 | mAP@Avg (%): 26.4 -> 26.9 | Improved |
| use full DELU instead of CO2-Net | CO2-Net · THUMOS14 | mAP@Avg (%): 44.6 -> 46.4 | Improved |
| use full DELU instead of UGCT | UGCT · THUMOS14 | mAP@Avg (%): 43.6 -> 46.4 | Improved |
| use DELU instead of ACM-Net | ACM-Net · ActivityNet1.2 | mAP@Avg (%): 26.5 -> 26.9 | Improved |
| add DELU to reduce action-background confusion | CO2-Net · THUMOS14 | AUROC (%): 83.96 -> 85.92 | Improved |
| add DELU to reduce background share in top 5% target CAS snippets | CO2-Net · THUMOS14 | background snippet share (%): 22.54% -> 20.92% | Improved |
Closest papers by meaning
- Uncertainty-Aware Dual-Evidential Learning for Weakly-Supervised Temporal Action Localization (TPAMI 2023 · similarity 0.96)
- Vectorized Evidential Learning for Weakly-Supervised Temporal Action Localization (TPAMI 2023 · similarity 0.94)
- Weakly-supervised Temporal Action Localization by Uncertainty Modeling (AAAI 2021 · similarity 0.92)
- Cascade Evidential Learning for Open-World Weakly-Supervised Temporal Action Localization (CVPR 2023 · similarity 0.90)
- Bridge the Gap: From Weak to Full Supervision for Temporal Action Localization with PseudoFormer (CVPR 2025 · similarity 0.88)
- 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.
