Leon Gatys is an AI researcher credited with Neural Style Transfer, an algorithm for creating artistic images. The method uses a deep neural network to separate and recombine the content and style of arbitrary images. Gatys developed it during his PhD under Prof. Matthias Bethge. It became one of the earliest demonstrations of the creative potential of generative AI.
In the research paper's examples, the content of a photograph of the Neckarfront in Tübingen, Germany is combined with the style of well-known artworks from different periods. The result keeps the global arrangement of the photograph while the colors and local structures come from the artwork. The artist Memo Akten recalls 2015 as a big year for deep learning, naming style transfer by Gatys alongside Google's Deep Dream. Style transfer is also cited among the technical innovations that shaped AI image generation, together with Alexander Mordvintsev's DeepDream and Ian Goodfellow's GANs.
