Thanks to Colormind who has figured out a way to produce beautiful color palettes with GAN training, we decided to take on similar approach for our solution too. GAN, also known as Generative Adversarial Network, is a form of supervised learning where you have a set of training data, input (A) and output (B) pairs, the machine learning algorithm will learn the mapping function from A to B.
We handpicked a range of colors and also beautiful colour palettes to train XTOPIA and the biggest takeaway from this exercise is the quality of our data set determines the quality of the output. Trash in trash out. Sure it sounds like a simple concept but a lot of hard work has been put in place to keep the quality of our training data within our expectation.
Why 25 colours? XTOPIA Website Builder allows user to pick up to 25 colours (tints and hues included) to design a website so a meaningful set of colors is needed to fit into the system.