What Is GAN? Generative Adversarial Networks (GAN) Explained

Key takeaways

A generative adversarial network (GAN) is a type of machine learning model designed to imitate the structure and function of a human brain.

• Two types of neural networks, generators and discriminators, make up a generative model.

• Three advantages of GANs include learning from unlabeled data, identifying anomalies in the modeled data, and creating realistic data samples.

• You can think of an artificial neural network (ANN) as a structure designed to imitate the functions of a human brain.

Learn more about GANs, including how generative models function, types of GAN models, and the pros and cons of this technology. If you’re ready to enhance your skill set in this field, enroll in the Deep Learning Specialization offered by DeepLearning.AI. In as little as three months, you can learn about artificial neural networks, PyTorch, computer vision, performance tuning, and more.

What is GAN technology?

GAN stands for generative adversarial network. It’s a type of machine learning model called a neural network, specially designed to imitate the structure and function of a human brain. For this reason, neural networks in machine learning are sometimes referred to as artificial neural networks (ANNs). This technology is the basis of deep learning, a subcategory of machine learning capable of recognizing complex patterns in varying data types such as images, sounds, and text. 

Learn more about real-world applications of GANs in this lecture from DeepLearning.AI's Generative Adversarial Networks (GANs) Specialization:

Human neurons vs. artificial neurons

Neurons in your brain are cells that use chemicals and electrical signals to send information between different parts of your brain and body. Neurons in machine learning models are modules of software that send information and computations to one another. 

Read more: 10 Machine Learning Algorithms to Know

Generative models

Generative models can generate new data samples by interpreting how data is placed in addition to what the data represents. In contrast, discriminative models focus on differentiating between existing data samples. 

Neural networks

Generative models are made up of two types of neural networks:

1. Generator: Generators are convolutional neural networks (CNN), a type of deep learning algorithm that can process an input image, differentiate between the objects within it, and assign importance to each one. These degrees of importance are known as weights. A generator network aims to create outputs that could be mistaken for real data. 

2. Discriminator: Discriminators are deconvolutional neural networks (DNN). These algorithms work in reverse to CNNs, aiming to identify the features of an input that were either missed by the CNN or convoluted with other signals. A discriminator network aims to identify which received output is artificial.

Adversarial setting

In the context of GANs, adversarial describes the training environment for each neural network (DNN and CNN).

During training, the generator and discriminator networks compete with one another in a bluffing game. The generator creates artificial data samples, such as fake images, to trick the discriminator into accepting them as authentic. In response, the discriminator attempts to identify which data samples are real images and which are not. They practice this game over and over, each one improving at their role each time. 

Generative vs. discriminative models

Suppose you provided this line of characters to a discriminative model and a generative model: 

^  ⌄ ^  ⌄ ^  ⌄

A generative model can predict how likely it is for a downward arrow to appear next in line. A discriminative model can decide which symbols are upward-facing and which are downward-facing. While the generative model also recognizes which arrows are facing in which direction, it takes its analysis further by assigning a probability to a sequence of symbols. Discriminative models instead focus on how likely the labels “upward” and “downward” are to apply to each symbol.

Types of GAN models

Researchers are still identifying new use cases for GANs and improving upon existing GAN techniques. Here are a couple of examples of different types of GANs:

• CycleGAN: Cycle generative adversarial networks focus on image-to-image translations. The training data set consists of two unpaired sets of data or groups of images with no labels or correspondences. The CycleGAN uses this information to learn how to transform images from one set into images that could pass for belonging to the other set. For example, suppose you provided a CycleGAN with two sets of images: one depicting house cats and one depicting tigers. The output might look like a realistic image of a house cat with tiger stripes. Or, the inverse might depict a house-cat-sized tiger.

• Super-resolution GANs: SRGANs are trained to increase image resolution by filling in details to blurry areas of an image. They accomplish this using perceptual loss function, a technique that measures the difference between the high-level perceptual features of two images. This technique allows for a low-resolution image to be upscaled to a high-resolution image.

Pros and cons of GANs

Discover more about machine learning

Subscribe to our weekly LinkedIn newsletter, Career Chat, for industry updates, tips, and trends. Then, explore free machine learning resources to optimize your professional growth:

• Watch on YouTube: Machine Learning in Real Life: From Spotify to Healthcare

• Earn a credential: 6 Machine Learning Certificates + How to Choose the Right One for You

• Explore a path: Machine Learning Career Paths: Explore Roles & Specializations

Whether you want to develop a new skill, get comfortable with an in-demand technology, or advance your abilities, keep growing with a Coursera Plus subscription. You’ll get access to over 10,000 flexible courses. 

Build job-ready skills with Coursera Plus

Start 7-day free trial

• 
• 
• 
• 

Start 7-day free trial