text-to-image diffusion models

Generative models that create images from text descriptions by gradually refining a noise-filled image through a series of steps. They leverage the diffusion process to achieve high-quality visual outputs based on textual input.

20 papers