Research map / Video / Video-language understanding
Vid2Seq: Large-Scale Pretraining of a Visual Language Model for Dense Video Captioning
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 |
|---|---|---|---|
| add pretraining on trimmed video-speech segments (ablation) | Vid2Seq · YouCook2 val | CIDEr: 18.0 -> 27.8 | Helped |
| pretrain on untrimmed videos instead of trimmed segments (ablation) | Vid2Seq · YouCook2 val | CIDEr: 27.8 -> 35.0 | Helped |
| add time tokens from speech sentence boundaries in pretraining (ablation) | Vid2Seq · YouCook2 val | CIDEr: 35.0 -> 47.1 | Helped |
| add denoising pretraining loss (ablation) | Vid2Seq (visual+speech) · YouCook2 val | CIDEr: 10.3 -> 47.1 | Helped |
| add speech input at finetuning to a visual-only pretrained model (ablation) | Vid2Seq generative-only pretraining · YouCook2 val | CIDEr: 25.3 -> 10.3 | Hurt |
| joint captioning with localization (vs localization-only) (ablation) | Vid2Seq pretrained · YouCook2 val | localization recall: 21.4 -> 27.8 | Helped |
| T5-Base instead of T5-Small (ablation) | Vid2Seq · YouCook2 val | CIDEr: 31.1 -> 47.1 | Helped |
| scale pretraining data 15K -> 15M videos (ablation) | Vid2Seq · YouCook2 val | CIDEr: 35.0 -> 47.1 | Helped |
| Vid2Seq vs PDVC/UEDVC | event localization · ActivityNet Captions val | localization recall: 59.0 vs 52.7 | Hurt |
Closest papers by meaning
- DIBS: Enhancing Dense Video Captioning with Unlabeled Videos via Pseudo Boundary Enrichment and Online Refinement (CVPR 2024 · similarity 0.91)
- Event-Equalized Dense Video Captioning (CVPR 2025 · similarity 0.89)
- Distilling Vision-Language Models on Millions of Videos (CVPR 2024 · similarity 0.89)
- VidChain: Chain-of-Tasks with Metric-based Direct Preference Optimization for Dense Video Captioning (AAAI 2025 · similarity 0.89)
- Seq2Time: Sequential Knowledge Transfer for Video LLM Temporal Grounding (CVPR 2025 · similarity 0.88)
- Video ReCap: Recursive Captioning of Hour-Long Videos (CVPR 2024 · similarity 0.88)
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.