The problem Monet Image Generator solves

1. A GAN architecture that captures special characteristics of one image collection, in our
   case Monet paintings, and translates these characteristics into the other image collection, all in the
   absence of any paired training examples.
2. It generates 7,000 to 10,000 Monet-style images which would trick the classifiers into believing that
   we have created a true monet using CycleGAN.
3. It builds a generator neural network responsible for creating the images, and a discriminator neural
   network which is trained to distinguish between real and generated images.
4. GANs help us create more realistic images which can further be used to generate new datasets for machine learning models where we have less data, as well as creating high resolution images from blurry low resolution images.

Challenges we ran into

CycleGAN was a new technology that we enquired while doing this project , thus understanding that in a short time was a challenge for us . But after looking into various resources we were able to learn about the metholodgy and working of the same.