Research map / Video / Video segmentation and tracking
Exploring Motion and Appearance Information for Temporal Sentence Grounding
Daizong Liu, Xiaoye Qu, Pan Zhou, Yang Liu
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 |
|---|---|---|---|
| Use MARN with motion and object features instead of the strongest proposal-based baseline CBLN | MARN · Charades-STA (official split) | R@1, IoU=0.7: 38.22 -> 43.09 | Improved |
| Use MARN with motion and object features instead of CBLN | MARN · TACoS (official split) | R@1, IoU=0.5: 27.65 -> 35.74 | Improved |
| Replace C3D motion features with I3D features | MARN · Charades-STA (official split) | R@1, IoU=0.7: 43.09 -> 44.80 | Improved |
| Replace C3D motion features with I3D features | MARN · TACoS (official split) | R@1, IoU=0.5: 35.74 -> 37.25 | Improved |
| Add the appearance branch (ablation) | MARN baseline · Charades-STA | R@1, IoU=0.7: 34.76 -> 37.20 | Improved |
| Add the motion encoder (ablation) | MARN with appearance branch · Charades-STA | R@1, IoU=0.7: 37.20 -> 38.91 | Improved |
| Add the motion branch (ablation) | MARN with appearance branch and motion encoder · Charades-STA | R@1, IoU=0.7: 38.91 -> 40.85 | Improved |
| Add the motion-appearance associating module (ablation) | MARN with appearance and motion branches · Charades-STA | R@1, IoU=0.7: 40.85 -> 43.09 | Improved |
| Remove global frame features (ablation) | MARN · Charades-STA | R@1, IoU=0.7: 43.09 -> 41.46 | Hurt |
| Remove position encoding (ablation) | MARN · Charades-STA | R@1, IoU=0.7: 43.09 -> 40.73 | Hurt |
| Use concatenation instead of attention for cross-modal interaction (ablation) | MARN reasoning branch · Charades-STA | R@1, IoU=0.7: 43.09 -> 41.62 | Hurt |
| Remove the spatio-temporal graph (ablation) | MARN reasoning branch · Charades-STA | R@1, IoU=0.7: 43.09 -> 41.38 | Hurt |
Closest papers by meaning
- Comprehensive Visual Grounding for Video Description (AAAI 2024 · similarity 0.88)
- Motion-Grounded Video Reasoning: Understanding and Perceiving Motion at Pixel Level (CVPR 2025 · similarity 0.88)
- Universal Video Temporal Grounding with Generative Multi-modal Large Language Models (NeurIPS 2025 · similarity 0.86)
- PixFoundation 2.0: Do Video Multi-Modal LLMs Use Motion in Visual Grounding? (NeurIPS 2026 · similarity 0.86)
- Empower Words: DualGround for Structured Phrase and Sentence-Level Temporal Grounding (NeurIPS 2025 · similarity 0.86)
- Decoupling Static and Hierarchical Motion Perception for Referring Video Segmentation (CVPR 2024 · 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.
