Research map / Video / Video-language understanding
SCANet: Scene Complexity Aware Network for Weakly-Supervised Video Moment Retrieval
Sunjae Yoon, Gwanhyeong Koo, Dahyun Kim, Chang D. Yoo
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 scene-complexity redundancy removal using nouns (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 27.32 -> 31.82 | Improved |
| use noun and verb together for redundancy removal instead of noun alone (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 31.82 -> 30.42 | Hurt |
| replace fixed proposal count with complexity-adaptive proposals (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 27.98 -> 31.82 | Improved |
| replace sliding window with complexity-adaptive proposals (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 26.51 -> 31.82 | Improved |
| use Flatten Gaussian mask instead of Gaussian mask (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 29.43 -> 31.82 | Improved |
| add hierarchical contrastive learning to cross-modal reconstruction (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 28.45 -> 30.32 | Improved |
| add loss calibration by scene complexity (ablation) | SCANet · ActivityNet Captions (val) | R@1, IoU=0.5: 30.32 -> 31.82 | Improved |
| use more than 15 retrieved videos as corpus-level hard negatives (ablation) | SCANet · ActivityNet Captions | R@5, IoU=0.1 | Hurt |
| SCANet vs CNM on ActivityNet R@1 IoU=0.5 | SCANet · ActivityNet Captions | R@1, IoU=0.5: 31.52 vs 33.33 (CNM) | No clear effect |
Closest papers by meaning
- VLANet: Video-Language Alignment Network for Weakly-Supervised Video Moment Retrieval (ECCV 2020 · similarity 0.91)
- Fewer Steps, Better Performance: Efficient Cross-Modal Clip Trimming for Video Moment Retrieval Using Language (AAAI 2024 · similarity 0.89)
- MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval (NeurIPS 2025 · similarity 0.89)
- Zero-shot Video Moment Retrieval via Off-the-shelf Multimodal Large Language Models (AAAI 2025 · similarity 0.88)
- Transferable Video Moment Localization by Moment-Guided Query Prompting (AAAI 2024 · similarity 0.87)
- Generative Video Diffusion for Unseen Novel Semantic Video Moment Retrieval (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.
