Research map / Video / Video generation
Preacher: Paper-to-Video Agentic System
Jingwei Liu, Ling Yang, Hao Luo, Fan Wang, Hongyan Li, Mengdi Wang
Read the paperCheck your idea against 137,217 papers
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 Preacher agentic pipeline instead of OpenAI-o3-mini + Sora | video generation · Paper-to-video benchmark (40 papers) | GPT accuracy: 4.33 -> 4.50 | Improved |
| use Preacher instead of OpenAI-o3-mini + Sora | video generation · Paper-to-video benchmark (40 papers) | Human accuracy: 3.88 -> 4.80 | Improved |
| use Preacher instead of OpenAI-o3-mini + Sora | video generation · Paper-to-video benchmark (40 papers) | CLIP: 0.31 -> 0.26 | Hurt |
| use Preacher domain-specific styles instead of Sora | video generation · Paper-to-video benchmark (40 papers) | AE: 5.31 -> 5.20 | Hurt |
| use Preacher key scene planning instead of GPT-4o direct planning | key scene planning · Key scene planning (40 papers) | accuracy: 4.05 -> 4.70 | Improved |
| use Preacher instead of DeepSeek-R1 planning | key scene planning · Key scene planning (40 papers) | compatibility: 3.70 -> 4.38 | Improved |
| use Preacher instead of o3-mini planning | key scene planning · Key scene planning (40 papers) | professionalism: 4.53 -> 4.63 | No clear effect |
| add progressive planning, reflection and other mechanisms to key scene planning (ablation) | Preacher · Key scene planning (40 papers) | accuracy | Improved |
| add multi-round reflection to key scene planning (ablation) | Preacher · Key scene planning (40 papers) | alignment | Hurt |
Closest papers by meaning
- VideoPhy: Evaluating Physical Commonsense for Video Generation (ICLR 2025 · similarity 0.83)
- ChronoMagic-Bench: A Benchmark for Metamorphic Evaluation of Text-to-Time-lapse Video Generation (NeurIPS 2024 · similarity 0.82)
- SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction (ICLR 2024 · similarity 0.80)
- Grid Diffusion Models for Text-to-Video Generation (CVPR 2024 · similarity 0.80)
- FreeNoise: Tuning-Free Longer Video Diffusion via Noise Rescheduling (ICLR 2024 · similarity 0.80)
- VideoMAR: Autoregressive Video Generation with Continuous Tokens (NeurIPS 2025 · similarity 0.80)
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.
