GPT-4 + Stable-Diffusion = ?: Enhancing Prompt Understanding of Text-to-Image Diffusion Models with Large Language Models
TL;DR: Text Prompt -> LLM -> Intermediate Representation (such as an image layout) -> Stable Diffusion -> Image. Recent advancements in text-to-image generation with diffusion models have yielded remarkable results…