GAN

Full Form: Generative Adversarial Network

Category: AI Architecture

📖 Definition

A GAN consists of two neural networks competing against each other: a generator that creates fake data and a discriminator that tries to distinguish real from fake. This competition improves both networks.

🔑 Key Points

💡 Why It Matters

GANs pioneered AI image generation, but have largely been superseded by diffusion models for most applications.