Research map / Vision-language and multimodal / Human and embodied modeling
DOC2PPT: Automatic Presentation Slides Generation from Scientific Documents
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 |
|---|---|---|---|
| Replace the flat progress tracker with hierarchical section-slide-object tracking (ablation) | DOC2PPT · DOC2PPT test set | structural LCS: 51.72% -> 64.15% | Helped |
| Add the paraphrasing loss (ablation) | Hierarchical PT without PAR, with TIM · DOC2PPT test set | ROUGE-SL: 26.52 -> 29.40 | Helped |
| Add the paraphrasing loss (ablation) | Hierarchical PT without PAR or TIM · DOC2PPT test set | TFR: 8.58 -> 9.23 | Helped |
| Add the text-figure matching loss (ablation) | Hierarchical PT without PAR · DOC2PPT test set | LC-FS F1: 20.99 -> 24.36 | Helped |
| Add the text-figure matching loss (ablation) | Hierarchical PT without PAR · DOC2PPT test set | LC-FS recall: 26.26 -> 25.31 | Hurt |
| Add the text-figure matching loss (ablation) | Hierarchical PT without PAR · DOC2PPT test set | ROUGE-SL: 24.93 -> 26.52 | Helped |
| Enable figure post-processing to remove low-relevance figures and add high-relevance ones (ablation) | Hierarchical PT with PAR and TIM · DOC2PPT test set | LC-FS F1: 24.36 -> 31.26 | Helped |
| Enable figure post-processing to remove low-relevance figures and add high-relevance ones (ablation) | Hierarchical PT with PAR and TIM · DOC2PPT test set | TFR: 11.82 -> 17.49 | Helped |
| Use learned layout prediction instead of template-based layout (ablation) | DOC2PPT · DOC2PPT test set | mIoU: 42.96 -> 50.82 | Helped |
| Condition the paraphraser on the object state (ablation) | DOC2PPT paraphraser · DOC2PPT test set | ROUGE-L: 31.95 -> 32.27 | Helped |
Closest papers by meaning
- AutoPresent: Designing Structured Visuals from Scratch (CVPR 2025 · similarity 0.88)
- Paper2Poster: Towards Multimodal Poster Automation from Scientific Papers (NeurIPS 2025 · similarity 0.84)
- PPT-Eval: A Benchmark for Computer-Use Agents on PowerPoint Tasks (ICML 2026 · similarity 0.83)
- BizGen: Advancing Article-level Visual Text Rendering for Infographics Generation (CVPR 2025 · similarity 0.81)
- Read, Parse, Describe: Unified Document Parsing with Visual Element Description Generation (NeurIPS 2026 · similarity 0.81)
- P2P: Automated Paper-to-Poster Generation and Fine-Grained Benchmark (ICLR 2026 · similarity 0.81)
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.