Research map / Video / Action recognition
Multi-Shot Temporal Event Localization: A Benchmark
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 |
|---|---|---|---|
| Add temporal aggregation to the vanilla method (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 16.1 -> 18.9 | Helped |
| Use multi-scale temporal aggregation rather than a single-scale module (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 17.5 -> 18.9 | Helped |
| Replace temporal aggregation with dilated 1D convolution (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 17.1 | Hurt |
| Replace temporal aggregation with deformable 1D convolution (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 17.0 | Hurt |
| Use a single temporal aggregation module with kernel 1×3 and W=3 instead of multiple scales (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 16.0 | Hurt |
| Use a single temporal aggregation module with kernel 3×3 and W=3 instead of multiple scales (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 16.8 | Hurt |
| Use a single temporal aggregation module with kernel 3×3 and W=6 instead of multiple scales (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 17.5 | Hurt |
| Use a single temporal aggregation module with kernel 3×3 and W=9 instead of multiple scales (ablation) | MUSES localization baseline · MUSES | mAP@IoU=0.5: 18.9 -> 17.2 | Hurt |
| Add temporal aggregation to the THUMOS14 localization method | THUMOS14 localization · THUMOS14 | mAP@IoU=0.5: +5.3% | Helped |
| Integrate temporal aggregation with BMN | BMN · ActivityNet-1.3 validation | mAP: 33.85 -> 33.99 | No clear effect |
Closest papers by meaning
- One-Shot Action Localization by Learning Sequence Matching Network (CVPR 2018 · similarity 0.89)
- HyrCap: Hybrid Rank-Calibration of Action Proposals for Temporal Event Understanding (NeurIPS 2026 · similarity 0.88)
- Test-Time Zero-Shot Temporal Action Localization (CVPR 2024 · similarity 0.87)
- Weakly-Supervised Temporal Action Localization by Inferring Salient Snippet-Feature (AAAI 2024 · similarity 0.87)
- PASS: Path-selective State Space Model for Event-based Recognition (NeurIPS 2025 · similarity 0.87)
- Realigning Confidence with Temporal Saliency Information for Point-Level Weakly-Supervised Temporal Action Localization (CVPR 2024 · similarity 0.86)
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.