Research map / Video / Video-language understanding
To Find Where You Talk: Temporal Sentence Localization in Video with Attention Based Location Regression
Read the paperCheck 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 global-view attention-based coordinate regression instead of candidate-clip matching | ABLR · ActivityNet Captions | mIoU: 0.2416 -> 0.3699 | Helped |
| add attention calibration loss (ablation) | ABLR with attended feature regression · ActivityNet Captions | mIoU: 0.3509 -> 0.3572 | Helped |
| add attention calibration loss (ablation) | ABLR with attention weight regression · ActivityNet Captions | mIoU: 0.3501 -> 0.3699 | Helped |
| add video Bi-LSTM context encoding (ablation) | ABLR with attended feature regression · ActivityNet Captions | mIoU: 0.3356 -> 0.3572 | Helped |
| add video Bi-LSTM context encoding (ablation) | ABLR with attention weight regression · ActivityNet Captions | mIoU: 0.2434 -> 0.3699 | Helped |
| retain sentence-level features and co-attention rather than a single Skip-thought vector (ablation) | ABLR with attention weight regression · ActivityNet Captions | R@1, IoU@0.5: 0.3094 -> 0.3679 | Helped |
| replace attention post-processing with coordinate regression (ablation) | ABLR · ActivityNet Captions | mIoU: 0.2917 -> 0.3699 | Helped |
| use attention weight regression rather than attended feature regression (ablation) | ABLR full · ActivityNet Captions | mIoU: 0.3572 -> 0.3699 | Helped |
| use attention weight regression rather than attended feature regression (ablation) | ABLR full · TACoS | R@1, IoU@0.5 | Hurt |
| use global-view regression rather than scan-and-localize methods | ABLR · ActivityNet Captions | localization time: 0.12s -> 0.02s | Helped |
| use global-view regression rather than scan-and-localize methods | ABLR · TACoS | localization time: 5.29s -> 0.15s | Helped |
Closest papers by meaning
- Unsupervised Domain Adaptative Temporal Sentence Localization with Mutual Information Maximization (AAAI 2024 · similarity 0.91)
- Transferable Video Moment Localization by Moment-Guided Query Prompting (AAAI 2024 · similarity 0.88)
- Omnipotent Distillation with LLMs for Weakly-Supervised Natural Language Video Localization: When Divergence Meets Consistency (AAAI 2024 · similarity 0.88)
- Localizing Events in Videos with Multimodal Queries (CVPR 2025 · similarity 0.87)
- Boundary-Aware Temporal Dynamic Pseudo-Supervision Pairs Generation for Zero-Shot Natural Language Video Localization (AAAI 2025 · similarity 0.87)
- Implicit Location-Caption Alignment via Complementary Masking for Weakly-Supervised Dense Video Captioning (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.