AI Dictionary › Fondamenti AI

Diffusion Model

Modello di diffusione

Diffusion models are a type of generative model that creates images, audio, or other data by starting from pure random noise and cleaning it up step by step. During training they learn the reverse process: take a real image, add noise progressively until it's unrecognizable, and the model learns to undo each step. When generating, it starts from a blob of noise and 'develops' it, like a photo in a darkroom, until a coherent image emerges. Guided by a text description, the model produces content matching the request.

Definition

Diffusion models are today the technology behind many high-quality image generators, having surpassed earlier approaches like GANs in fidelity and stability. Their iterative nature makes them a bit slower but enables fine control, targeted edits, and remarkably realistic, varied results.

Related terms

• Generative Adversarial Network (GAN)
• Autoencoder
• Encoder-Decoder
• Computer Vision
• Deep Learning

More in Fondamenti AI

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.