AI Dictionary › Fondamenti AI

Generative Adversarial Network (GAN)

Rete generativa avversaria (GAN)

A GAN, generative adversarial network, is a system of two neural networks competing against each other. The generator tries to create fake but believable data, such as nonexistent faces; the discriminator tries to tell real data from generated. It's like a duel between a forger and a detective: the forger improves to deceive, the detective sharpens their eye to expose it, and this rivalry drives both to ever higher levels. As training proceeds, the generator produces results so convincing the discriminator struggles to catch the fakery.

Definition

GANs marked a breakthrough in realistic image generation and style transfer, but are known for being unstable and hard to train. In demanding uses they've been partly supplanted by diffusion models, though they remain historically important and useful where fast, sharp single-pass generation is needed.

Related terms

• Diffusion Model
• Autoencoder
• Neural Network
• 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.