For the world’s greatest artists, creating a masterpiece takes years. This is made up of practice, experimentation, doodling through earlier works, blood-sweat-and-tears, and perhaps cutting off the odd ear. Now a computer program could accurately re-create any work by any artist within one hour.
This claim comes from German researchers, who released a science paper which explains how a computer algorithm can produce images, lifting from the styles of the world’s greatest artists. The exact basis of the algorithm has not been disclosed; however, it layers a “convolutional neural network” which means, The Guardian explains, it “uses object recognition to recreate the foundation image (which can be anything) in the style of a piece of specific art.”
To illustrate the point the paper contains images which the authors say have been developed from running the algorithm. These pictures are placed side-by-side with pictures painted by such luminaries as Vincent Van Gogh and Pablo Picasso.
Speaking with The Washington Post, lead author of the paper, Leon Gatys, a PhD student at the University of Tuebingen stated: “The key finding of this paper is that the representations of content and style in the convolutional neural network are separable. That is, we can manipulate both representations independently to produce new, perceptually meaningful images.”
The paper has been published in the e-journal arXiv and titled “A Neural Algorithm of Artistic Style.”
