Research map / Video / Action recognition
Action Unit Memory Network for Weakly Supervised Temporal Action Localization
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 |
|---|---|---|---|
| Use AUMN instead of TSCN | Weakly supervised temporal action localization · THUMOS14 test | average mAP@IoU (0.1:0.1:0.5): 47.0 -> 52.1 | Helped |
| Use AUMN instead of Bas-Net | Weakly supervised temporal action localization · ActivityNet1.2 validation | average mAP@IoU: 24.3 -> 25.5 | Helped |
| Use AUMN instead of A2CL-PT | Weakly supervised temporal action localization · ActivityNet1.3 validation | average mAP@IoU: 22.5 -> 23.5 | Helped |
| Add sparsity loss (ablation) | AUMN · THUMOS14 | average mAP@IoU (0.1:0.1:0.5): 43.2 -> 49.2 | Helped |
| Add diversity loss to the sparsity-loss model (ablation) | AUMN · THUMOS14 | average mAP@IoU (0.1:0.1:0.5): 49.2 -> 50.2 | Helped |
| Add homogeneity loss to the sparsity-loss model (ablation) | AUMN · THUMOS14 | average mAP@IoU (0.1:0.1:0.5): 49.2 -> 49.8 | Helped |
| Add diversity and homogeneity losses to sparsity loss (ablation) | AUMN · THUMOS14 | average mAP@IoU (0.1:0.1:0.5): 49.2 -> 51.4 | Helped |
| Add self-attention to the model with all three auxiliary losses (ablation) | AUMN · THUMOS14 | average mAP@IoU (0.1:0.1:0.5): 51.4 -> 52.1 | Helped |
| Add diversity and homogeneity losses together rather than using either alone with sparsity (ablation) | AUMN · THUMOS14 | mAP@IoU 0.5: 28.8; 29.5; 29.4 -> 31.8 | Helped |
Closest papers by meaning
- Weakly Supervised Action Localization by Sparse Temporal Pooling Network (CVPR 2018 · similarity 0.93)
- Completeness Modeling and Context Separation for Weakly Supervised Temporal Action Localization (CVPR 2019 · similarity 0.92)
- Relational Prototypical Network for Weakly Supervised Temporal Action Localization (AAAI 2020 · similarity 0.92)
- Temporal Action Localization in Untrimmed Videos via Multi-Stage CNNs (CVPR 2016 · similarity 0.91)
- Background Suppression Network for Weakly-Supervised Temporal Action Localization (AAAI 2020 · similarity 0.91)
- Weakly-Supervised Temporal Action Localization by Inferring Salient Snippet-Feature (AAAI 2024 · 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.