A spectacular advancement in the field of artistic analysis has been achieved thanks to the use of artificial intelligence. Indeed, British and American researchers have recently employed an AI method to study in depth an iconic painting by Raphael, the Madonna della Rosa. This project revealed a somewhat surprising detail: the face of Saint Joseph, visible in the artwork, appears not to have been painted by the master himself. This finding raises new questions about the authenticity and the artistic techniques of the time.
The AI analysis technique
The researchers developed a customized algorithm specifically designed to analyze Raphael’s style based on verified works. This innovation combines a pre-trained architecture known as ResNet50, developed by Microsoft, along with a traditional machine learning method called Support Vector Machine (SVM). Dr. Hassan Ugail, a mathematician at the University of Bradford, explains that “this method allows for a meticulous examination of brush strokes, color palettes, and shading techniques used by Raphael“, thus providing a very detailed understanding of his style.
Disturbing results
During their analysis of the Madonna della Rosa, the researchers confirmed that the faces of the Virgin Mary, the Child Jesus, and Saint John the Baptist indeed came from the illustrious painter. However, the face of Saint Joseph drew attention for unexpected reasons. The test results showed a significant divergence from the rest of the artwork. Dr. Ugail specified: “When we examined the painting as a whole, the results were inconclusive. This led us to test each part individually, revealing that, very likely, the face of Joseph is not by the hand of Raphael.”
An echo of history
This revelation was in some ways an echo of doubts long expressed by art historians. Since the mid-19th century, the face of Saint Joseph has sparked interest among critics who have always considered it executed in a less complete manner than the other characters in the painting. The present analysis only corroborates these age-old queries, thus providing new perspectives to an exciting debate on the validity of artwork attributions.
A human-machine collaboration
It is worth noting that, although AI has been a valuable tool in this research, experts insist that human expertise remains essential in the authentication process of a work of art. Dr. Ugail states: “AI does not replace the work of researchers. The authentication of a work involves a multitude of analyses, from its provenance to the materials used, not to mention the overall condition of the piece.” Thus, AI is now seen as an effective complement to the expertise of specialists, thereby enhancing the collaborative approach in the field of art.
Future perspectives
The discovery made by the research team also opens new avenues for artistic analysis. The application of cutting-edge technologies, such as artificial intelligence, could transform our understanding of the world’s artistic heritage. By using methods capable of detecting elements that the human eye cannot perceive, such as microscopic details, experts are now able to deepen their investigations into other masterpieces, thus shedding light on artistic mysteries long buried.
This marriage between artificial intelligence and human expertise represents a promising advancement, transforming the way we study and understand the history of art, while unveiling new facets of the works of the Great Masters.
