Paper2Poster: A Multi-Modal Poster Generator from Papers
What is Paper2Poster?
Paper2Poster is an open-source multi-agent system jointly developed by institutions including the University of Waterloo in Canada, the National University of Singapore, and the University of Oxford. Its PosterAgent multi-agent pipeline can automatically generate multimodal posters from scientific papers through a parser, planner, and painter–commenter loop.

Features of Paper2Poster
Long-text compression: Compresses multi-page scientific papers into a single-page poster while retaining the core content of the paper.
Multimodal content processing: Extracts multimodal content such as text, charts, and images from the paper and integrates them into the poster in a reasonable way.
Layout optimization: Generates aesthetically pleasing and logically clear poster layouts, ensuring that content is reasonably distributed within the limited space.
Visual quality enhancement: Optimizes the visual effects of the poster based on a visual feedback mechanism to improve its readability.
Evaluation and optimization: Evaluates the poster's ability to convey core content through PaperQuiz and optimizes it based on feedback.
Technical Principles of Paper2Poster
The core of Paper2Poster is PosterAgent, which consists of three components:
Parser: Converts PDFs to Markdown format and then generates a structured asset library in JSON format.
Planner: Aligns text and visual elements to generate a binary tree layout, maintaining reading order and spatial balance.
Painter–Commenter (Painter–Commenter): Generates the specific content of each panel and optimizes panel layouts based on a visual feedback mechanism.
How to Use Paper2Poster
Installation: Supports local deployment (via vLLM) or API access (such as GPT-4o). The installation steps include installing dependency packages and configuring environment variables.
Quick start: Place the paper's PDF file in the designated directory and run a specific command to generate the poster.
Advantages of Paper2Poster
Efficiency: Can complete poster generation within minutes, significantly saving time and effort compared to traditional manual methods.
Low cost: Generation cost is as low as $0.005, greatly lowering the barrier to academic communication.
Open source: The open-source weights have been released on the HuggingFace platform, allowing users to use it for free and optimize it according to their needs.
Application Scenarios of Paper2Poster
Academic conferences: Helps researchers quickly convert papers into posters for conference presentations.
Academic reports: Serves as auxiliary material for academic reports, helping audiences better understand research content.
Research achievement presentation: Displays the latest research achievements in research institutions or laboratories.
Education field: Teachers generate posters for teaching purposes, helping students understand complex academic concepts more intuitively.
Project Resources
Project website:https://paper2poster.github.io/
GitHub repository:https://github.com/Paper2Poster/Paper2Poster
HuggingFace model library:https://huggingface.co/datasets/Paper2Poster/Paper2Poster
arXiv technical paper:https://arxiv.org/pdf/2505.21497
Script Planner Assistant▸
AI Apps
AI News 