Artificial Intelligence at the service of Art and the conservation of artistic heritage

The National Research Council through the Institute of Cultural Heritage Sciences (Cnr-Ispc) has announced that it has achieved significant progress in the study and conservation of works of art. Researchers were able to precisely and rapidly examine and assemble enormous volumes of data generated by X-ray spectroscopic techniques, thanks to the introduction of a revolutionary method based on artificial intelligence.

The new methodology, which could change Heritage Science forever, was successfully tested on two fragments of the Baronci Altarpiece by Raffaello Sanzio, preserved in the Capodimonte Museum in Naples. The research, published in ‘Science Advances‘, marks a significant turning point in the field of non-invasive investigations applied to pictorial works. The deep learning algorithm, trained on a large synthetic dataset of XRF spectra, accurately analyzes the data, offering a new perspective on the chemical composition and distribution of pigments without the common artifacts of previous analyses.

The method proposed by the Cnr-Ispc not only improves the quality of the analysis but also the reliability of the interpreted data. This allows you to get hitherto inaccessible details on the painting techniques and the state of conservation of the workscontributing to a more effective preservation of our cultural heritage, but what turns out to be truly revolutionary is the exclusive use of synthetic data for training the artificial intelligence model without the need for real samples, opening up new possibilities for the application of AI in the field of Heritage Science.