Chinese scientists restore more than 100 ancient paintings with AI technology

Chinese researchers are revolutionising the restoration of ancient paintings through cutting-edge artificial intelligence, breathing new life into fading masterpieces and safeguarding cultural heritage for future generations. This is reported by CCTV+, a partner of TV BRICS.

At Zhejiang University’s Laboratory of Art and Archaeology Image, experts are using AI-driven restoration technology to digitally revive the original colours of historic artworks.

By integrating accumulated knowledge, methodologies, and previous research on art restoration into a single AI model, experts have developed an invaluable AI-powered restoration assistant.

Traditionally, the restoration of ancient paintings relied heavily on the expertise of conservators, who manually analysed pigments and performed digital simulations. With AI-driven models, however, the process has been transformed, as reported by the source.

Advanced algorithms now analyse pigment composition, the painting’s historical context, and the artist’s distinctive style to generate an accurate colour restoration. The AI then produces a visual simulation, which conservators meticulously verify and refine in the laboratory.

The source also detailed how the research team replicates historical pigments by applying them to silk and paper using traditional techniques. This hands-on approach ensures that digital restorations remain as true as possible to the artwork’s original appearance.

So far, the team has restored over 100 ancient paintings through AI technology. Additionally, researchers are compiling extensive data on colour compositions, materials, and artistic styles to further train and refine their AI model.

Their long-term objective is to build a comprehensive database encompassing the entire history of Chinese painting, allowing for even more precise restorations in the future.