Diffusion Model

Definition

A generative AI model that creates data by learning to reverse a gradual noising process. During training, it learns to remove noise step by step. During generation, it starts with pure noise and progressively refines it into coherent output. Powers image generators like DALL-E 3, Stable Diffusion, and Midjourney.

How It Works

Diffusion models are the technology behind AI image generators like Stable Diffusion, DALL-E 3, and Midjourney. They work by learning to reverse a process of gradually adding noise to images. During training, the model sees images at various stages of noise corruption and learns how to remove the noise step by step. During generation, it starts with pure random noise and iteratively refines it into a coherent image.

The math is elegant: define a forward process that slowly destroys an image by adding Gaussian noise over many steps. Then train a neural network to predict and remove that noise at each step. Run the reverse process, and noise becomes art. The quality comes from doing this in many small steps rather than one big leap — each step only needs to make a small improvement.

Diffusion models displaced GANs (Generative Adversarial Networks) as the dominant image generation approach around 2022. They're more stable to train, produce higher-quality results, and offer better control over the output. They've also been adapted beyond images — for video generation (Sora), music, 3D models, and even protein structure prediction. The main downside is speed: all those iterative denoising steps make generation slower than single-pass models.