AI Glossary Term
What does “Diffusion Model” mean?
A generative architecture that learns to reverse noise to synthesize images, audio, or other content.
How Diffusion Model works
Example
An image generator is trained by taking real pictures, adding noise step by step, and learning to reverse that process. At creation time it starts from random noise and removes it gradually, steered by the text prompt, until a picture appears.
Why it matters
Diffusion models power many current image, audio, and video generators. They also raise questions about training data rights and about fake imagery.
Common questions
Why start from noise?
The model learned to remove noise at each step, so repeating that from pure noise produces new content that follows patterns from training data.
Are diffusion models only for images?
No. They are also used for audio, video, 3D shapes, and molecules.
Why can generation be slow?
It takes many denoising steps, so researchers work on methods that need fewer steps.