AI ART Discover · Learn · Create
AI Art > Glossary > GAN (Generative Adversarial Network)
Browse Full AI Glossary
G AI Glossary Term • Term 86 of 224
Plain English Definition

What does “GAN (Generative Adversarial Network)” mean?

Core Definition

A generative setup where a generator and discriminator train against each other.

Operational Mechanism

How GAN (Generative Adversarial Network) Works

1

Generator makes a fake

2

Discriminator compares fake vs real

3

Discriminator says which is which

4

Both improve, repeat

5

back to the start.

Real-World Scenario

Practical Example

In an early GAN face generator, one network invents faces while another tries to tell them from photos. As the detector gets better, the generator must produce more convincing faces to fool it.

Core Significance

Why It Matters

GANs produced some of the first very realistic synthetic images and helped start public debate about deepfakes. Diffusion models have since overtaken them in many uses.

Expert Q&A

Common Questions About GAN (Generative Adversarial Network)

Q: Who invented GANs?

Ian Goodfellow and colleagues introduced them in 2014.

Q: Why are GANs hard to train?

The two networks must stay balanced. If one overpowers the other, training can stall or the generator may repeat a narrow set of outputs, called mode collapse.

Q: Are GANs still used?

Yes, in areas such as image enhancement and synthetic data, though diffusion models are now more common for general image generation.