Over 130 years ago, Norwegian painter Edvard Munch took a life-changing evening stroll. As the sun dipped below the horizon, the crimson clouds inspired him to create The Scream (1893), a masterpiece that has captivated audiences for generations.

However, time and exposure have taken their toll on Munch's iconic work. There are four versions of The Scream—two paintings and two pastel drawings—all of which have suffered from fading and degradation after more than a century. Adding to the artwork’s woes, the 1910 version fell victim to theft in 2004. Although it was returned and recovered in 2006, new moisture damage marred its lower-left side. The vibrant colors that Munch once envisioned now seem increasingly elusive.

Inevitably, all artworks face the same destiny of age. A consortium of 12 organizations has come together under the European Union-backed PERCEIVE project to develop a solution: authentic color reconstruction using advanced AI tools. Launched in the past year, PERCEIVE represents a global effort to construct a service-oriented AI architecture and toolkit accessible to both experts and the public. Notable institutions include the MUNCH Museum in Oslo, Norway; the Art Institute of Chicago; the Victoria & Albert Museum in London; and the National Archaeological Museum of Naples. are collaborating on this three-year project. Tech pioneers such as Fraunhofer-IGD and imki lend their expertise to train and build AI infrastructure. The project's goal is to enhance the capabilities of conservators and curators to digitally reconstruct the original colors of five key groups of art: statues, paintings and works on paper, textiles, photographs, and augmented reality artworks.

Representing the MUNCH Museum, Irina Crina Anca Sandu, the project lead, unveiled “The Scream Time Machine.” This interactive program digitally reconstructs Edvard Munch’s iconic painting, “The Scream,” step by step—from its 1893 inception to speculative versions in 2093 and beyond. As PERCEIVE’s funding extends until 2026, researchers aspire to transform these prototypes into accessible tools. A color knowledge repository, a light-damage estimator, or a web-based color prediction service, etc. their impact on art preservation promises to be profound.

On the other hand, digital reproduction or manipulation of artworks raises philosophical questions. Veteran conservation scientist William Wei noted the importance of understanding what can be done with an artifact because restoring it means changing its certain features. 

Take “The Scream” as an example, Sandu emphasized the need to bring in the context when interpreting data. She added that conservators must reflect on Munch's inspiration from nature and his struggles with mental health, which led him to paint The Scream. In Sandu’s view, Munch’s turmoil transcends time. “The Scream” is humanity’s universal heritage—a message resonating across generations. Together with AI technologies and prototyping, we carry this legacy into the future, honoring both science and soul. 

Sayart / Amia Nguyen, amyngwyen13@gmail.com