PosterGen: Aesthetic-Aware Paper-to-Poster Generation via Multi-Agent LLMs
Zhilin Zhang, Xiang Zhang, Jiaqi Wei, Yiwei Xu, Chenyu You
2025-08-26
Summary
This paper introduces a new system called PosterGen that automatically creates research posters from scientific papers using the power of large language models and a team of specialized AI 'agents'.
What's the problem?
Creating a good research poster is time-consuming, and existing automated tools don't do a great job with the visual design aspects. They often produce posters that look messy or don't follow good design principles, meaning researchers still have to spend a lot of time fixing them manually.
What's the solution?
PosterGen works like a team of professional poster designers. It has four different AI agents that work together: one to understand the paper and pick out key information, one to arrange that information on the poster, one to choose colors and fonts, and finally, one to put everything together into a finished poster. This collaborative approach aims to create posters that are both accurate and visually appealing.
Why it matters?
This work is important because it can save researchers a significant amount of time and effort when preparing for conferences. By automatically generating high-quality posters, PosterGen allows scientists to focus more on their research and less on the tedious task of poster design, and the new way to evaluate poster quality using AI could help improve future poster generation systems.
Abstract
Multi-agent systems built upon large language models (LLMs) have demonstrated remarkable capabilities in tackling complex compositional tasks. In this work, we apply this paradigm to the paper-to-poster generation problem, a practical yet time-consuming process faced by researchers preparing for conferences. While recent approaches have attempted to automate this task, most neglect core design and aesthetic principles, resulting in posters that require substantial manual refinement. To address these design limitations, we propose PosterGen, a multi-agent framework that mirrors the workflow of professional poster designers. It consists of four collaborative specialized agents: (1) Parser and Curator agents extract content from the paper and organize storyboard; (2) Layout agent maps the content into a coherent spatial layout; (3) Stylist agents apply visual design elements such as color and typography; and (4) Renderer composes the final poster. Together, these agents produce posters that are both semantically grounded and visually appealing. To evaluate design quality, we introduce a vision-language model (VLM)-based rubric that measures layout balance, readability, and aesthetic coherence. Experimental results show that PosterGen consistently matches in content fidelity, and significantly outperforms existing methods in visual designs, generating posters that are presentation-ready with minimal human refinements.