Text2Poster: Laying out Stylized Texts on Retrieved Images

Abstract

Poster generation is a significant task for a wide range of applications, which is often time-consuming and requires lots of manual editing and artistic experience. In this paper, we propose a novel data-driven framework, called \textit{Text2Poster}, to automatically generate visually-effective posters from textual information. Imitating the process of manual poster editing, our framework leverages a large-scale pretrained visual-textual model to retrieve background images from given texts, lays out the texts on the images iteratively by cascaded auto-encoders, and finally, stylizes the texts by a matching-based method. We learn the modules of the framework by weakly- and self-supervised learning strategies, mitigating the demand for labeled data. Both objective and subjective experiments demonstrate that our Text2Poster outperforms state-of-the-art methods, including academic research and commercial software, on the quality of generated posters.

5 pages, Accepted to ICASSP 2022

References in corpus (8)

• Adam: A Method for Stochastic Optimization
• Representation Learning with Contrastive Predictive Coding
• Zero-Shot Text-to-Image Generation
• Learning Visual Importance for Graphic Designs and Data Visualizations
• PP-OCR: A Practical Ultra Lightweight OCR System
• WenLan: Bridging Vision and Language by Large-Scale Multi-Modal Pre-Training
• Constrained Graphic Layout Generation via Latent Optimization
• Bilinear Representation for Language-based Image Editing Using Conditional Generative Adversarial Networks