🎨 PosterVerse

[AAAI 2026 Oral] PosterVerse: A Full-Workflow Framework for Commercial-Grade Poster Generation with HTML-Based Scalable Typography

Official Hugging Face release of PosterVerse models and dataset.


🌟 Highlights

  • We propose PosterVerse, a full-workflow framework integrating:

    • Blueprint Creation
    • Graphical Background Generation
    • Unified Layout-text Rendering

    enabling commercial-grade poster generation with sophisticated layouts and text-rich design.

  • We introduce PosterDNA, the first commercial-grade, text-dense poster dataset with fine-grained HTML-based specifications, supporting modular training and evaluation.

  • PosterVerse supports end-to-end poster generation from text prompts.

  • The framework produces visually appealing, layout-consistent, and text-accurate posters suitable for real-world commercial design scenarios.


🧠 Model Zoo

The PosterVerse pipeline consists of multiple stages:

1. Blueprint Creation

  • Repo: PosterVerse/Blueprint_Creation

2. Unified Layout-Text Rendering

  • Repo: PosterVerse/Unified_Layout-Text_Rendering

3. Graphical Background Generation

  • Repo: PosterVerse/lora_model

📦 Dataset

PosterDNA

  • File: PosterVerse/posterdna.zip

PosterDNA is a commercial-grade poster dataset with:

  • Text-dense layouts
  • Fine-grained HTML annotations
  • High-quality design samples

⚠️ Note:

  • The PosterDNA dataset can only be used for non-commercial research purposes. For scholar or organization who wants to use the PosterDNA dataset, you can apply through either of the following two options:

    Option A: Apply Online Submit your application through our online platform: 👉 Apply Here

    Option B: Apply via Email Please first fill in this Application Form and sign the Legal Commitment and email them to us (eelwjin@scut.edu.cn). When submitting the application form to us, please list or attached 1-2 of your publications in the recent 6 years to indicate that you (or your team) do research in the related research fields of poster generation, layout design, font generation, and so on.

  • We will give you the decompression password after your application has been received and approved.

  • All users must follow all use conditions; otherwise, the authorization will be revoked.


✒️Citation

If you find PosterVerse helpful, please consider giving this repo a ❤ and citing:

@inproceedings{Liu2026posterverse,
      title={PosterVerse: A Full-Workflow Framework for Commercial-Grade Poster Generation with HTML-Based Scalable Typography}, 
      author={Junle Liu, Peirong Zhang, Yuyi Zhang, Pengyu Yan, Hui Zhou, Xinyue Zhou, Fengjun Guo, Lianwen Jin},
      journal={Proceedings of the AAAI Conference on Artificial Intelligence},
      year={2026},
}

Thanks for your support!


Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support