FACES technology will help scientists in the study of works of art and establish their authorship along with radiography and other instruments. The project will be presented in April next year at the New York Museum Frick Collection.
Next year on the museum’s website Frick Collection will be available a tool that uses face recognition software of unknown models in portraits. Users will be able to upload photos and use Art and Computerized Evaluation Systems (FACES) to search for the database. The project was launched by scientists from the University of California at Riverside (UCR) with the financial support of the National Foundation for the Support of the Humanities.
FACES uses the same technology that governments use to track down terrorists. The tool will be available on the website of the Reference Arts Library, which is part of the New York Museum. “I hope that scientists will use this tool along with radiography and other technologies,” says Konrad Rudolf, professor of art at UCR. He emphasizes that the task of this technology is not to replace the expert’s view, but to help him in the research.
Unlike the earlier version of FACES, which measured the distance between anthropometric points on the face, “version 2.0” uses deep neural networks that take into account a larger number of points and more subtly and in detail analyze complex, nonlinear visual information. “True, artistic liberties, characteristic of portraits created by painters, will never allow us to use universal technology for all works,” Rudolf notes. At the moment, the museum has already explored portraits of many famous personalities, including William Shakespeare and Robert Devereaux, a favorite of Elizabeth I. But, according to Rudolph, the potential of software is not limited to portraits. It can be used to establish authorship of works in other spheres of art. The project will be presented in April 2018 at a symposium in the Frick Collection.