Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.
I'm working on Pix2Pix an image-to-image translation GAN, and I noticed that there is an adversarial loss implemented using BCE, and a L1 loss implemented using MAE. I know L1 loss represents the difference between the predicted image and actual image, but I am not sure what does the GAN adversarial loss represent?
This is the official definition
The adversarial loss influences whether the generator model can output images that are plausible in the target domain
but the meaning is tough to understand. Is it representing the difference between the predicted probability distribution and actual probability distribution?
The adversarial loss in a GAN represents the amount of information that the generator is able to trick the discriminator into believing is true.
"The adversarial loss influences whether the generator model can output images that are plausible in the target domain"$\rightarrow$ means that the generator model should output images that look like they could plausibly belong to the target domain. This is important because if the generated images look too fake, it will be difficult for the discriminator to learn to correctly classify them. It's the data that determines whether and how it improves.
In addition, the adversarial loss is based on the generator model's experience of its successes and failures in the data.