Research map / Video / Video-language understanding
End-to-end Multi-modal Video Temporal Grounding
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 |
|---|---|---|---|
| average outputs of extra modality streams (flow, depth) without fusion modules | LGI baseline · Charades-STA | R@0.5: 59.16 -> 57.39 | Hurt |
| add co-attentional transformer fusion (ablation) | three-stream DRFT · Charades-STA | R@0.5: 61.05 -> 63.03 | Helped |
| use RGB as the common modality instead of flow (ablation) | three-stream DRFT · Charades-STA | R@0.5: 61.20 -> 63.03 | Helped |
| add shared co-attention block for RGB (ablation) | three-stream DRFT · Charades-STA | R@0.5: 62.61 -> 63.03 | Helped |
| add learnable dynamic fusion weights (ablation) | three-stream DRFT · Charades-STA | R@0.5: 61.65 -> 63.03 | Helped |
| add intra-modal contrastive loss (ablation) | three-stream DRFT · Charades-STA | R@0.5: 61.87 -> 63.03 | Helped |
| add learnable fusion weights (ablation) | three-stream DRFT · ActivityNet Captions | R@0.5: 44.42 -> 45.72 | Helped |
| add co-attentional transformer (ablation) | three-stream DRFT · ActivityNet Captions | R@0.5: 43.83 -> 45.72 | Helped |
| add intra-modal contrastive loss (ablation) | three-stream DRFT · ActivityNet Captions | R@0.5: 44.50 -> 45.72 | Helped |
| add depth and flow modalities with proposed modules (single to three-stream) | DRFT · Charades-STA | R@0.5: 60.79 -> 63.03 | Helped |
Closest papers by meaning
- Local-Global Multi-Modal Distillation for Weakly-Supervised Temporal Video Grounding (AAAI 2024 · similarity 0.92)
- SnAG: Scalable and Accurate Video Grounding (CVPR 2024 · similarity 0.89)
- SimVG: A Simple Framework for Visual Grounding with Decoupled Multi-modal Fusion (NeurIPS 2024 · similarity 0.87)
- Contextual Self-paced Learning for Weakly Supervised Spatio-Temporal Video Grounding (ICLR 2025 · similarity 0.87)
- PixFoundation 2.0: Do Video Multi-Modal LLMs Use Motion in Visual Grounding? (NeurIPS 2026 · similarity 0.87)
- MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models (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.