PosterO: Structuring Layout Trees to Enable Language Models in Generalized Content-Aware Layout Generation

TL;DR

PosterO introduces a novel layout-centric method leveraging large language models to generate diverse, content-aware poster layouts by structuring design elements as trees, achieving state-of-the-art results and supporting generalized design scenarios.

Contribution

It proposes a new approach that structures layout trees in SVG language and uses LLMs for generalized, content-aware poster generation, addressing limitations of prior image-centric methods.

Findings

Achieves state-of-the-art performance on multiple benchmarks.

Successfully generates diverse, visually appealing poster layouts.

Demonstrates effectiveness in generalized, multi-purpose poster design scenarios.

Abstract

In poster design, content-aware layout generation is crucial for automatically arranging visual-textual elements on the given image. With limited training data, existing work focused on image-centric enhancement. However, this neglects the diversity of layouts and fails to cope with shape-variant elements or diverse design intents in generalized settings. To this end, we proposed a layout-centric approach that leverages layout knowledge implicit in large language models (LLMs) to create posters for omnifarious purposes, hence the name PosterO. Specifically, it structures layouts from datasets as trees in SVG language by universal shape, design intent vectorization, and hierarchical node representation. Then, it applies LLMs during inference to predict new layout trees by in-context learning with intent-aligned example selection. After layout trees are generated, we can seamlessly realize them into poster designs by editing the chat with LLMs. Extensive experimental results have demonstrated that PosterO can generate visually appealing layouts for given images, achieving new state-of-the-art performance across various benchmarks. To further explore PosterO's abilities under the generalized settings, we built PStylish7, the first dataset with multi-purpose posters and various-shaped elements, further offering a challenging test for advanced research.