GlyphDraw2: Automatic Generation of Complex Glyph Posters with Diffusion Models and Large Language Models

• DOI: 10.48550/arxiv.2407.02252
• Publication Date: 2024-07-02
• Authors: Jian Ma; Yong-lin Deng; Chen Chen; H. Lu; Zhenyu Yang
• Keywords: Glyph (data visualization)

Posters play a crucial role in marketing and advertising, contributing significantly to industrial design by enhancing visual communication brand visibility. With recent advances controllable text-to-image diffusion models, more concise research is now focusing on rendering text within synthetic images. Despite improvements accuracy, the field of end-to-end poster generation remains underexplored. This complex task involves striking balance between accuracy automated layout produce high-resolution images with variable aspect ratios. To tackle this challenge, we propose an framework employing triple cross-attention mechanism rooted align learning, designed create precise detailed contextual backgrounds. Additionally, introduce dataset that exceeds 1024 pixels image resolution. Our approach leverages SDXL architecture. Extensive experiments validate ability our method generate featuring intricate contextually rich Codes will be available at https://github.com/OPPO-Mente-Lab/GlyphDraw2.