A masterpiece hanging in the Museo Nacional del Prado in Madrid has long sparked debates over whether it was the work of Raphael. But a group of researchers now claims to have finally solved the mystery through the use of an artificial intelligence algorithm.
The Madonna della Rosa (Madonna of the Rose) depicts Mary, Joseph, and the baby Jesus, along with an infant version of John the Baptist. Until the 19th century, the painting was attributed to the Italian Renaissance painter Raffaello Sanzio da Urbino, more often known as Raphael. Then doubts were raised over the Joseph figure “looking like an afterthought” and whether Raphael had painted the lower section.
The museum’s website page for the oil painting solely credits it to Raphael.
According to a new research paper published on December 21 in the journal Heritage Science, analysis of the painting using an AI algorithm with an accuracy of 98 percent found that the painting was entirely made by the Italian artist. But it “raised questions about whether Raphael indeed painted the face of Joseph in the painting.”
The researchers, led by University of Bradford visual computer professor Hussan Ugail, noted that the AI analysis supported earlier work by art historians who had “previously questioned the full attribution of this painting to Raphael alone, suggesting that his associate, Giulio Romano, might have had a hand in it.”
University of Bradford emeritus professor of molecular spectroscopy Howell Edwards, who co-authored the paper, told the Guardian: “The AI program analysis of our work has demonstrated conclusively that whereas the three figures of the Madonna, [Jesus] and St John the Baptist are unequivocally by Raphael, that of St Joseph is not, and has been painted by someone else.”
In January, Ugail was part of a team of researchers who used AI-assisted computer-based facial recognition on a painting known as the de Brécy Tondo to also help determine it was a work by Raphael. The research team found that the faces of the Madonna and child in the de Brécy Tondo were identical to ones in the Raphael altarpiece Sistine Madonna. Then another study called into question the results of that research, and museum experts raised questions about the methodology.
Ugail told the Guardian that he knows “nothing about art”, and that the reception to his work from art historians can be frosty. “I think there is fear and they also think we are naive, that we don’t know what we are doing,” he said.
While there is ongoing concern over how the use of AI will eliminate the work of human beings, the research team emphasized in the conclusion of their Heritage Science paper that AI could become a useful resource for art historians and collectors as a supplementary tool for verifying paintings alongside “existing methods such as scholarly analysis, spectroscopic imaging, and dating techniques.”
“As advances continue to be made in machine learning and image processing technologies, this method has the potential to become part of an array of tools for artwork analysis and verification,” the paper said. “It can operate in conjunction with other methods currently in use, including in-depth scrutiny by art historians and various advanced imaging techniques, thus contributing to a more thorough and dependable framework for artwork authentication and analysis.”