Diffusion Models
Generative models that learn to create data by reversing a gradual noising process.
Diffusion models add noise to data incrementally, then learn to denoise. State-of-the-art for image generation (Stable Diffusion, DALL-E). Produce high-quality, diverse outputs.
Examples
- Stable Diffusion for text-to-image
- DALL-E 3 for creative imagery
- Imagen for photorealistic generation