Russia Uses AI to Digitally Restore Ancient Frescoes
A Neural Network Recreates the Original Colors and Details of Centuries-Old Artwork
In Russia’s Pskov Technopark “Quantorium”, researchers are using artificial intelligence to digitally reconstruct the original appearance of the frescoes in the Snetogorsky Monastery, whose colors and details have faded over the centuries. The project was described by Victoria Zaretskaya, Deputy Director for Project Management at the technopark, during an interview on the “Minkult” program on Silver Rain Radio in Pskov.
The AI system processes photographs of the frescoes in their current condition and recreates what experts believe to be their original colors and artistic form. Zaretskaya emphasized that the project involves digital restoration only, without any physical interference with the artwork.
Restoring the Original Image
Specialists use modern photographs of the frescoes, whose pigments have changed under the influence of time, humidity, and light. The neural network analyzes these images and generates reconstructions showing how the murals might have looked when first created. For restorers, this digital reconstruction helps visualize the lost details and artistic intent behind the centuries-old paintings.
The initiative is part of the “Tekhnogorye” project, which focuses on developing technological tools for studying and preserving cultural heritage at the Snetogorsky Monastery — combining history, art, and AI to bring Russia’s medieval masterpieces closer to their original form.
