Research map / Video / Video-language understanding
Aligning Moments in Time using Video Queries
Yogesh Kumar, Uday Agarwal, Manish Gupta, Anand Mishra
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 post-fusion alignment (ablation) | MATR (no pretraining) · ActivityNet-VRL | R@1: 50.2 -> 52.9 | Improved |
| add pre-fusion alignment (ablation) | MATR (no pretraining) · ActivityNet-VRL | R@1: 50.2 -> 51.2 | Improved |
| add both pre- and post-fusion alignment (ablation) | MATR (no pretraining) · SportsMoments | R@1: 46.6 -> 52.7 | Improved |
| add both pre- and post-fusion alignment (ablation) | MATR (no pretraining) · ActivityNet-VRL | mIoU: 49.7 -> 53.2 | Improved |
| add self-supervised pre-training without augmentation (ablation) | MATR · ActivityNet-VRL | R@1: 54.8 -> 55.6 | Improved |
| add augmentation to pre-training (ablation) | MATR · ActivityNet-VRL | R@1: 55.6 -> 57.1 | Improved |
| add augmentation to pre-training (ablation) | MATR · SportsMoments | R@1: 53.6 -> 56.5 | Improved |
| use prediction heads instead of reading from post-fusion alignment matrix (ablation) | MATR · ActivityNet-VRL | mIoU: 38.3 -> 56.8 | Improved |
| use soft-DTW alignment instead of no alignment loss (ablation) | MATR · SportsMoments | R@1: 48.6 -> 56.5 | Improved |
| use soft-DTW alignment instead of TCC (ablation) | MATR · ActivityNet-VRL | R@1: 52.7 -> 57.1 | Improved |
| use soft-DTW instead of Drop-DTW (ablation) | MATR · SportsMoments | R@1: 54.7 -> 56.5 | Improved |
| finetune on video queries versus trained-from-scratch MATR | MATR · ActivityNet-VRL | R@1: 54.8 -> 57.1 | Improved |
Closest papers by meaning
- When One Moment Isn't Enough: Multi-Moment Retrieval with Cross-Moment Interactions (NeurIPS 2025 · similarity 0.90)
- Transferable Video Moment Localization by Moment-Guided Query Prompting (AAAI 2024 · similarity 0.90)
- MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval (NeurIPS 2025 · similarity 0.89)
- Fewer Steps, Better Performance: Efficient Cross-Modal Clip Trimming for Video Moment Retrieval Using Language (AAAI 2024 · similarity 0.89)
- Anchor-Aware Similarity Cohesion in Target Frames Enables Predicting Temporal Moment Boundaries in 2D (CVPR 2025 · similarity 0.89)
- 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.
