Generative Diffusion

Generative Diffusion refers to a class of machine learning models that generate data (such as images, audio, or text) by gradually transforming random noise into a coherent output through a process known as diffusion. In the context of image generation, the model starts with random noise and iteratively refines it by following learned patterns to produce a high-quality image.