Quick Introduction to GANs

“(GANs), and the variations that are now being proposed is the most interesting idea in the last 10 years in ML, in my opinion.” Yann Lecun

GANs have been getting a lot of attention since their introduction by Ian Goodfellow, the results are impressive and also promising. From generating images to increasing resolutions to image translation, GANs have not failed to surprise everyone.

So why are GANs so helpful? It is often that the probability distribution of the data is very complicated and difficult to infer, but GANs can learn to generate samples from the nasty probabilistic distribution of the data without even us dealing with it. Isn’t it nice!

Don’t know much about GANs? I know that’s why you are here. In this series we will go through the basic concepts of GANs, how do they work, drawbacks and also implement some.

• Toy-GAN
• DCGANs