Diffusion Model
A type of AI that creates images by starting with pure visual noise, like the static on an old television, and gradually cleaning it up into a clear picture that matches your description. This step-by-step denoising approach is called a diffusion model, and it powers many of the popular image generators.
A diffusion model is the engine behind a lot of AI image generation. The name comes from the way it works, which is more surprising than you might expect. It does not paint a picture stroke by stroke. It begins with a field of random noise and slowly removes that noise, step by step, until an image emerges.
Here is the rough idea. During training, the model is shown countless images with noise gradually added until each becomes static, and it learns to reverse that process. When you ask for a picture, it runs that learning in the other direction: it starts from noise and, guided by your words, refines it into something that matches. It is a little like watching a photograph slowly develop, except the model decides what should appear as it goes.
This approach turned out to produce strikingly good results, which is why tools that make images from a written description often rely on it. When you use a text-to-image generator, there is a good chance a diffusion model is doing the work behind the scenes.
You do not need to understand the maths to make use of such tools. The value of knowing the term is recognising it when it appears in product descriptions or articles, and understanding that “diffusion” refers to this noise-to-image method rather than anything you have to manage yourself.