Diffusion Models are a type of machine learning model, mainly used to generate new data, like images, sounds, or even text.
They work in two main steps:
Adding Noise (Forward Process): They start with real data (like a real image) and gradually add random noise to it, step by step, until the image becomes pure noise.
Removing Noise (Reverse Process): Then, they learn how to reverse this process — meaning, they start with random noise and gradually remove the noise to create a brand-new, realistic-looking sample (like a new image you've never seen before).