Researchers Develop AI Tool to Rapidly Restore Age-Damaged Artworks within Hours

Oil paintings often suffer from wear, leaving cracks and discoloration, making restoration a lengthy process reserved for valuable works. A new technique utilizing artificial intelligence and digital tools offers a rapid solution, achieving restorations in hours. This method, pioneered by MIT researcher Alex Kachkine, involves creating a digital reconstruction of the damaged artwork, which is then printed on a transparent polymer sheet and overlaid on the original. In a demonstration, Kachkine restored a late 15th-century painting attributed to the Master of the Prado Adoration, which had extensive damage. Using a digital mask and color-matching techniques, he enhanced 57,314 colors in the restoration process. The method, although effective, raises ethical concerns about the use of overlays and authenticity. Experts believe it could enable galleries to display otherwise hidden, damaged artworks, expanding public access to art while still considering the implications of using such techniques on valuable pieces.

Source link