Can an algorithm confirm or debunk the authorship of a work of art? Is artificial intelligence entitled to attribute or deny authorship? In the University of Bradford (England) They believe it, but in the scientific community it is viewed with great caution, if not disbelief. Last January, the BBC revealed that a work had been attributed to Rafael thanks to facial recognition technology. A team from the universities of Nottingham and Bradford used the technology to examine a painting known as Brécy Tondo. It is in a collection owned by Cheshire businessman George Lester Winward and some experts believed it to be a Victorian copy. And it was also discovered that the faces were identical to those on an altarpiece by Raphael “Madonna Sistine”. The similarity between the Madonnas is 97%, while comparing the child in both works showed a similarity of 86%. It was concluded that it was very likely that the paintings were by the same artist.
However, in September we learned that there was a war between artificial intelligences: two teams were fighting over the authorship of a painting by Rafael. A new study by the company Art Recognition came to the opposite conclusion about the “Brécy Tondo”. Soon after, a gallery (Cartwright Hall in Bradford, UK) turned a deaf ear and went a step further, exhibiting the AI-authenticated painting for the first time. This caught the attention of Carina Popovici, from Art Recognition. “At first I was excited and saw it as something positive in the artistic community. However, when doubts arose about its authenticity, we at Art Recognition decided to conduct our own AI-based authenticity analysis,” he explained on social media. The algorithms used here by the specialist resulted in exactly that opposite result: An 85 percent chance that the painting is not the work of Raphael.
Now an AI algorithm developed by Hasan Ugail, Professor of Visual Computing at the University of Bradford, returns to the fray with another painting by Raphael, in this case from the Prado collection. Accordingly, “The Guardian” publishes information “Holy Family with San Juanito” (known as “La Vigen de la Rosa”), from the Spanish Museum, was not entirely created by the master: most of the painting is by Raphael, but José’s face is by another hand. The discovery article appeared today in the journal Heritage Science.
Howell EdwardsEmeritus Professor of Molecular Spectroscopy at Bradford and co-author of the paper, said: “The AI program analysis of our work has shown conclusively that although the three figures of the Virgin [Jesús] and John the Baptist are clearly by Raphael, but the image of Saint Joseph is not and was painted by someone else.
Consulted by ABC, Prado does not want to comment on this On the subject. On the museum’s website, the poster accompanying the work states: “In 1633, Vicente Carducho mentioned in his workshop a painting by Raphael in poor condition that this could be this one, suggesting that it was made from a painting It was probably transferred to canvas in the first half of the 19th century, changing its dimensions and adding the lower band with the rose that ironically gives the work its name. The lack of original support and the Surface wear They make it difficult to date the painting, which may have been painted around 1517, although there is no information about its commission. There is no doubt about the attribution to Raphaelalthough the traditional interpretation of the painter’s Roman works as puzzles in which different members of the workshop would intervene in one part or another has led some specialists to consider it Participation of the “Bottega” in the execution of anecdotal parts of the characters.” At first glance, the figure of Joseph on the left side of the canvas appears significantly darker and seemingly less precisely painted.
