Research map / Video / Video-language understanding
Cascaded Prediction Network via Segment Tree for Temporal Video Grounding
Check 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 a cross-gate module instead of 1D convolution or 1D max-pooling for tree initialization (ablation) | CPN · ActivityNet Caption, Charades-STA, and TACoS | grounding performance | Helped |
| Add signal-loss supervision from inside target segments (ablation) | CPN · ActivityNet Caption, Charades-STA, and TACoS | grounding performance | Helped |
| Add message passing to the initialized segment tree (ablation) | CPN · ActivityNet Caption, Charades-STA, and TACoS | grounding performance | Helped |
| Set signal-loss balance factor λ to 6 (ablation) | CPN · ActivityNet Caption and Charades-STA | grounding performance | Helped |
| Set signal-loss balance factor λ to 8 (ablation) | CPN · TACoS | grounding performance | Helped |
| Increase the signal-loss weight from too small a value toward the dataset-specific optimum (ablation) | CPN · ActivityNet Caption, Charades-STA, and TACoS | grounding performance | Helped |
| Increase the signal-loss weight beyond the dataset-specific optimum (ablation) | CPN · ActivityNet Caption, Charades-STA, and TACoS | grounding performance | Hurt |
| Use fine-tuned rather than non-fine-tuned visual features | CPN · Charades-STA | grounding performance | Helped |
| Use 3D ConvNet rather than 2D ConvNet visual features | CPN · Charades-STA | grounding performance | Helped |
Closest papers by meaning
- Proposal-Free Video Grounding with Contextual Pyramid Network (AAAI 2021 · similarity 0.90)
- Comprehensive Visual Grounding for Video Description (AAAI 2024 · similarity 0.87)
- Universal Video Temporal Grounding with Generative Multi-modal Large Language Models (NeurIPS 2025 · similarity 0.86)
- Empower Words: DualGround for Structured Phrase and Sentence-Level Temporal Grounding (NeurIPS 2025 · similarity 0.86)
- VideoGrounding-DINO: Towards Open-Vocabulary Spatio-Temporal Video Grounding (CVPR 2024 · similarity 0.86)
- Temporal Sentence Grounding with Relevance Feedback in Videos (NeurIPS 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.