Gustav Klimt’s faculty paintings were stolen by the Nazis and destroyed by fire in the last days of World War II. With only black and white photos of the remaining works, it is unclear how the paintings were colored.
Based on a large dataset of Klimt’s works, archival data, and academic research, Google Arts and Culture used artificial intelligence (AI) to provide the best possible estimate of what the original paintings would look like.
Dr. Franz Smola, curator of the Belvedere Museum in Vienna was surprised with the result because they were able to color the places that they didn’t even know. With machine learning, they had good assumptions that Klimt used certain colors.
Klimt’s “Faculty Pictures” were representations of the faculties of philosophy, medicine and law on behalf of the Austrian Ministry of Culture of the University of Vienna. Klimt had to rent a new studio to work on the paintings, each of which were over 13 feet (about four meters) tall. He completed the collection in the early 20th century.
The university finally rejected the paintings which, instead of depicting a world inclined to science and reason, showed it to be agitated by sex, death and chaos. The works were sold privately before they were stolen by the Nazis in 1938 and taken to Immendorf Castle for safekeeping.
Immendorf Castle is presumably burned down on May 8, 1945 by the retreating German Wehrmacht and the “Faculty Pictures” and many other works by Klimt, are said to have been destroyed.
The machine learning project is part of a bold presentation of the artist’s work, curated by Smola, titled Klimt vs Klimt: The Man of Contradictions, which went online today.
Google Arts and Culture worked with 30 partners and institutions to showcase hundreds of paintings, drawings, letters and more, many of which were digitized for the first time.
Contains interactive essays on dozens of works and topics in Klimt’s paintings, hundreds of high-resolution images and an augmented reality pocket gallery with 63 Klimt masterpieces, including “College Paintings”.
Source link
