Cycle-GAN-Artworks

Welcome to computer generated artworks. In this project I just trained a Cycle-Generative Adversarial Nerwork or Cycle-GAN model on Google Colab and it took more than 4 hours to train. I scraped images from Bing and built my own dataset. Then I trained my Cycle-GAN model using those images. Here I sraped images with keyword Waterfall and Famous Abstract Art from Bing.

The Model

When training a Cycle-GAN there are four neural networks that need to be trained:

1. A generator that generates pictures of abstract paintings (abstract painting generator).
2. A generator that generates pictures of waterfalls (waterfall image generator).
3. A discriminator that can tell the difference between real and fake abstract paintings (abstract painting discriminator).
4. A discriminator that can tell the difference between real and fake pictures of waterfalls (waterfall image discriminator).

Using this model I've generated 85 images of size 256 x 256 pixels and made an art gallery website of computer generated images. This website is fully responsive and both desktop and mobile views are supported.

Code

Please check this Colab Notebook for the entire code.