Colormind can generate completely *unique* color schemes using a deep neural net.
Designers often use other sources of media as inspiration - film, photography, classical art and architecture. Colormind does this in a single click, by training on separate datasets of millions of images and videos.
Colormind can generate colors with your input - give it your brand's colors and it will fill out the rest of the palette in an intelligent way. It can learn from other media to offer different color options.
One of the biggest advances in deep learning came in 2016 in the form of GANs (Generative Adversarial Networks). This is a novel architecture that solves the "averaging problem" of regular deep learning neural nets.