Research map / Video / Video-language understanding
Weakly-Supervised Video Moment Retrieval via Semantic Completion Network
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 SCN instead of the strongest listed baseline WS-DEC | Weakly supervised moment retrieval · ActivityCaptions | R@1, IoU=0.1: 62.71 -> 71.48 | Helped |
| Use SCN instead of QSPN | Weakly supervised moment retrieval · ActivityCaptions | R@1, IoU=0.3: 52.12 -> 47.23 | Hurt |
| Use SCN instead of QSPN | Weakly supervised moment retrieval · ActivityCaptions | R@5, IoU=0.1: 77.72 -> 90.88 | Helped |
| Use SCN instead of QSPN | Weakly supervised moment retrieval · Charades-STA | R@1, IoU=0.3: 54.70 -> 42.96 | Hurt |
| Use SCN instead of QSPN | Weakly supervised moment retrieval · Charades-STA | R@5, IoU=0.3: 95.60 -> 95.56 | No clear effect |
| Add random proposal exploration during selection (ablation) | SCN · ActivityCaptions and Charades-STA | moment retrieval performance | Helped |
| Use reward feedback rather than one-hot labels for rank loss (ablation) | SCN · ActivityCaptions and Charades-STA | training convergence and moment retrieval performance | Helped |
| Mask important query words for semantic completion (ablation) | SCN · ActivityCaptions and Charades-STA | moment retrieval performance | Helped |
| Share parameters between proposal generation and semantic completion (ablation) | SCN · ActivityCaptions and Charades-STA | moment retrieval performance | Helped |
Closest papers by meaning
- Fewer Steps, Better Performance: Efficient Cross-Modal Clip Trimming for Video Moment Retrieval Using Language (AAAI 2024 · similarity 0.88)
- Generative Video Diffusion for Unseen Novel Semantic Video Moment Retrieval (AAAI 2025 · similarity 0.88)
- MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval (NeurIPS 2025 · similarity 0.87)
- Universal Video Temporal Grounding with Generative Multi-modal Large Language Models (NeurIPS 2025 · similarity 0.86)
- Zero-shot Video Moment Retrieval via Off-the-shelf Multimodal Large Language Models (AAAI 2025 · similarity 0.86)
- Anchor-Aware Similarity Cohesion in Target Frames Enables Predicting Temporal Moment Boundaries in 2D (CVPR 2025 · 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.