Microsoft has made their own AI-powered image generator, and it’s pretty meh

Everyone is doing it, which means Microsoft at least needs to try. In this case it is creating amazing images using Generative Adversarial Networks, a popular form of AI which uses two competing AI networks to create realistic images.

In this case, using the power of Azure and Kubernettes, Microsoft used the huge database of artwork from the Met to create realistic art that does not exist.

Microsoft writes:

We created an online experience called Gen Studio where users can explore generated, dreamlike images created by AI. These images are created using a Generated Adversarial Network (GAN), which can randomly sample from the space of possible artworks. We then take this trained GAN and invert it to create an algorithm capable of finding the closest match, in the GANs ‘mind’, to the real artwork in The Met. This allows us to not just create random works, but to interpolate between real artworks in the collection. We present two simple ways of exploring the GANs underlying 140-dimensional space of art.

Our first visualization lets users explore a two-dimensional slice of the vast “latent” space of the GAN. Users can move throughout this space and see how the GAN’s dreams change as they bump into real pieces in The Met’s collection. Our second visualization gives the user precise control of how to blend different works together into a larger work.

Unfortunately, unlike pictures of face and cats, I suspect the pictures of breast-plate styled handbags are rather unlikely to go viral.

Hear Microsoft talk about the project in their video below, and try to create your own artwork here.

Some links in the article may not be viewable as you are using an AdBlocker. Please add us to your whitelist to enable the website to function properly.