When you purchase through links on our site, we may earn an affiliate commission.Heres how it works.
GANs have several applications across different industries.
They can do this either through text-based prompts, or by modifying existing content.
In addition to creation, GAN can also be used to edit images.
One web link generates new data by taking an input data sample and modifying it as much as possible.
The other internet tries to predict whether the generated data output belongs to the original dataset.
The two neural networks that make up a GAN are referred to as the generator and the discriminator.
Generators act as the creative force behind the GAN.
It takes a random collection of numbers that initially hold no meaning.
This could be a realistic image, a snippet of music, or even a piece of text.
This data pipe plays a role similar to that of a discerning critic.
The discriminators job is to analyze both and determine whether they are real or fake.
How do GANs work?
The generator produces samples, and the discriminator evaluates them.
If the discriminator can quickly recognize the fake data that the generator produces, the generator suffers a penalty.
Another crucial aspect of GANs is a technique called backpropagation.
This is the process that powers how the generator learns from the discriminators feedback.
Backpropagation essentially allows the errors identified by the discriminator to be propagated backward through the generators layers.
Based on these errors, the weights and biases in the generators data pipe are adjusted.
This in turn helps the generator produce more realistic data points in the next iteration.
The training process is over once the discriminator can no longer recognize synthesized data.
A Vanilla GAN is the simplest form of a GAN.
Then theres the Conditional GAN (cGAN).
Similarly, if you wanted to generate pictures of dogs you could use a second GAN.
Researchers have also crafted various kinds of GAN architectures to generate music that captures the essence of human-like compositions.
