Research map / Video / Video-language understanding
Cross-Sentence Temporal and Semantic Relations in Video Activity Localisation
Jiabo Huang, Yang Liu, Shaogang Gong, Hailin Jin
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 temporal order constraint L_TMP (ablation) | CRM · ActivityNet-Captions test | temporal consistency accuracy: 45.02 -> 70.82 | Improved |
| add temporal order constraint L_TMP (ablation) | CRM · Charades-STA test | temporal consistency accuracy: 73.91 -> 74.65 | Improved |
| add semantic pairing constraint (ablation) | CRM · ActivityNet-Captions test | semantic consistency recall IoU=0.5: 57.84 -> 71.30 | Improved |
| add semantic pairing constraint (ablation) | CRM · Charades-STA test | semantic consistency recall IoU=0.5: 31.01 -> 51.33 | Improved |
| add temporal constraint to BCE baseline (ablation) | BCE base model · ActivityNet-Captions val_1 | recall | Improved |
| add semantic constraint to BCE baseline (ablation) | BCE base model · ActivityNet-Captions val_1 | recall | Improved |
| add both temporal and semantic constraints to BCE baseline (ablation) | BCE base model · ActivityNet-Captions val_1 | recall | Improved |
| stack more self-attention and cross-attention units (ablation) | CRM · ActivityNet-Captions val_1 | recall | Hurt |
| use cross-sentence relations versus single-sentence moment-based methods | CRM · ActivityNet-Captions val_1 | recall IoU@0.5: 30.73 -> 31.67 | Improved |
Closest papers by meaning
- Semantic Proposal for Activity Localization in Videos via Sentence Query (AAAI 2019 · similarity 0.91)
- Phrase-Level Temporal Relationship Mining for Temporal Sentence Localization (AAAI 2023 · similarity 0.90)
- Video Activity Localisation with Uncertainties in Temporal Boundary (ECCV 2022 · similarity 0.90)
- Bilateral Relation Distillation for Weakly Supervised Temporal Action Localization (TPAMI 2023 · similarity 0.89)
- METAL: Minimum Effort Temporal Activity Localization in Untrimmed Videos (CVPR 2020 · similarity 0.89)
- To Find Where You Talk: Temporal Sentence Localization in Video with Attention Based Location Regression (AAAI 2019 · similarity 0.89)
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.
