PosterGen: Aesthetic-Aware Paper-to-Poster Generation via Multi-Agent LLMs

2025-08-23

AI Summary

Academic conference poster generation often suffers from poor aesthetic quality, necessitating extensive manual refinement. To address this, we propose the first multi-agent collaborative framework explicitly designed for poster aesthetics. Our approach decomposes the task into four sequential stages—content parsing, layout planning, stylistic design, and rendering synthesis—and introduces a vision-language model (VLM) to establish a quantitative, interpretable aesthetic evaluation metric. Unlike existing end-to-end methods, our framework explicitly emulates professional design workflows by orchestrating stage-wise, large language model (LLM)-driven agent collaboration. Experiments demonstrate that our method matches state-of-the-art approaches in content fidelity while significantly improving visual quality—particularly in balance, visual hierarchy, and stylistic consistency. Moreover, it reduces manual editing time by 62%, enabling production-ready, presentation-grade posters with minimal human intervention.

Technology Category

• Multiagent Systems: Adversarial Agents
• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

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.

Problem

• Automating paper-to-poster generation while ensuring aesthetic quality
• Addressing neglect of design principles in automated poster creation
• Reducing manual refinement needs in research poster production

Innovation

• Multi-agent LLM framework automates poster design
• Specialized agents handle content, layout, styling
• Vision-language model evaluates aesthetic quality automatically