A digital artist has developed an Artificial Intelligence system that creates a realistic version of the faces of the protagonists of the most famous paintings
Technology and art are now two interconnected worlds. More and more artists are exploiting the potential of new technologies – from artificial intelligence to image processing systems – to create truly unique works of art. A few weeks ago, for example, a 10-second video sold at Christie’s auction for a staggering $6.6 billion. The short film is the work of Beeple, a famous Californian digital artist who has put together and animated more than 5000 PC images over the last 13 years. Another amazing application of new technology to the art world comes from Denis Shirayev, a digital artist who uses Artificial Intelligence to bring his artwork to life. In particular, Denis has developed a machine-learning algorithm that has allowed him to “create” a realistic version of the faces of the protagonists of seven internationally famous works of art, such as Botticelli’s Venus.
Artificial Intelligence to recreate the faces of works of art
Shirayev’s algorithm exploits artificial neural networks (ANN), i.e. mathematical/computational models inspired by the structure of biological neural networks. An ANN consists of a group of information interconnections made up of artificial neurons and processes using a computational connectionism approach. The advantage is that it is an adaptive system capable of changing its structure according to the external or internal information flowing through it during the learning phase. Thanks to this peculiarity of artificial neural networks, Denis Shirayev was able to analyze and map the entire surface of the selected pictures in a short time. Then, using image manipulation software, he reconstructed the subjects painted on canvas over 100 years ago in 4K and 60FPS. The end result is truly sharp, high-resolution images that show the “real face” of the subjects immortalized by the most famous painters in the history of art.
