LLM4GEN: Leveraging Semantic Representation of LLMs for Text-to-Image Generation
This addresses challenges in text-to-image generation for users dealing with dense prompts, offering a plug-and-play solution with strong performance gains.
The paper tackles the problem of text-to-image generation with complex prompts by proposing LLM4GEN, a framework that integrates LLM features into diffusion models, resulting in significant improvements such as 9.69% and 12.90% increases in color alignment for SD1.5 and SDXL on T2I-CompBench.
Diffusion models have exhibited substantial success in text-to-image generation. However, they often encounter challenges when dealing with complex and dense prompts involving multiple objects, attribute binding, and long descriptions. In this paper, we propose a novel framework called \textbf{LLM4GEN}, which enhances the semantic understanding of text-to-image diffusion models by leveraging the representation of Large Language Models (LLMs). It can be seamlessly incorporated into various diffusion models as a plug-and-play component. A specially designed Cross-Adapter Module (CAM) integrates the original text features of text-to-image models with LLM features, thereby enhancing text-to-image generation. Additionally, to facilitate and correct entity-attribute relationships in text prompts, we develop an entity-guided regularization loss to further improve generation performance. We also introduce DensePrompts, which contains $7,000$ dense prompts to provide a comprehensive evaluation for the text-to-image generation task. Experiments indicate that LLM4GEN significantly improves the semantic alignment of SD1.5 and SDXL, demonstrating increases of 9.69\% and 12.90\% in color on T2I-CompBench, respectively. Moreover, it surpasses existing models in terms of sample quality, image-text alignment, and human evaluation.