That Algorithm Is Biting My Style: Imitating the Masters, Artificial Intelligence Divides Style from …

by Ben M. Harvey An artist’s visual style is a tenuous thing. By looking at tendencies for particular colors, shapes, and patterns of brush strokes, an experienced viewer can immediately identify an unfamiliar Van Gogh or Cezanne—regardless of subject matter. But what distinguishes an artist’s visual style from the content they represent in a particular work? Scientists working in artificial intelligence are figuring that out. In a recent paper, Leon Gatys and colleagues in Tübingen, Germany, attempt to answer this question by replicating visual style using computer algorithms. They used Deep Convolutional Neural Networks to separate an artwork’s style—that is, its identifying aesthetic characteristics—from its content. In doing so, they were seemingly able to identify the singular style in a given artwork and replicate it across images with new subject…


Link to Full Article: That Algorithm Is Biting My Style: Imitating the Masters, Artificial Intelligence Divides Style from …