55_auto_x2_colored_toned_photos_v2_light_ai(3).jpg

Snapshot of a machine learning art. ​

Wiki photo database tagged with painting styles is used to train a generative model of art. This is a snapshot in the midst of the learning process. The model is trained to convert random samples from a Gaussian distribution to photos of art. The model’s goal is to “fool" an art critic. The critic is simultaneously trained to distinguish actual art from the generated images.