Computer Generated 'Rembrandt' Painting Unveiled in Amsterdam
A new “Rembrandt” painting unveiled in Amsterdam in early April is making headlines around the world. However, the authentic-looking masterpiece is not the work of the 17th-century Dutch artist. It is a brand new painting that uses technology to mimic Rembrandt’s technique so perfectly, that it could easily be mistaken for one created by the great artist himself.
The clever forgery that took 18 months to complete is the result of a collaboration between Netherland’s ING Bank, tech giant Microsoft, local advertising agency J. Walter Thompson, and art experts from Delft University of Technology (TU Delft), The Mauritshuis and Museum Het Rembrandthuis. Called the “Next Rembrandt,” it uses data and technology to resurrect one of the greatest artists of all times.
The team began by collecting data from the Dutch artist’s 346 paintings to help them emulate his technique, choice of color, structure, texture, and topic, as closely as possible. They then used facial recognition software and a unique deep learning algorithm to analyze the individual features of his style — things like painting strokes, the artist’s angle of choice, etc.
During their analysis, the researchers were able to collect a significant amount of data on Rembrandt's typical portraits of Caucasian males with black suits, white collars, and whiskers. More data meant the computer software could mimic the artist’s style better, which is why the team settled on creating a portrait.
Once the subject had been determined, a proprietary software system and facial recognition algorithm gathered information about Rembrandt's style based on his use of geometry, composition, and choice of paint colors. The data was then used to replicate the artist’s style and generate the facial features for the “Next Rembrandt.”
Once ready, the individual elements were put together to form the face and the bust in the same proportions as the original paintings created by the Dutch artist. The team then used technology to add depth and texture to the 2D image. Once ready, a 3D printer was used to bring the “Next Rembrandt,” – A portrait of a white male between 30 and 40 years old, wearing black clothes, a white collar, and hat — to life. Comprising of 148 million pixels and 13 layers of UV ink, the “painting” is a clever forgery that could easily be mistaken for an original Rembrandt — At least to the untrained eye.
J. Walter Thompson’s executive creative director, Bas Korsten, says the goal of the project was to start a discussion about how data and technology could become an essential part of the art world. While the project did receive some criticism from a few experts, many art historians, including Rembrandt experts from The Mauritshuis in The Hague (the museum where Rembrandt’s last painting is exhibited) and Museum Het Rembrandthuis in Amsterdam, were incredibly supportive. They even acted as advisors to ensure that the “Next Rembrandt” would be a masterpiece — One that even the famous Dutch artist would be proud of.
Born in Amsterdam in 1606, Rembrandt Harmenszoon van Rijn is one of the world’s most renowned artists. The prolific painter was famous for his portraits thanks to his ability to capture realistic emotions. Unfortunately, like many artists of his time, Rembrandt's talents were not recognized during this lifetime. The artist died penniless in 1669, after suffering through many years of hardship.
Resources: livescience.com, metmuseum.org, thenextrembrandt.com
Reading Comprehension (3 questions)
- What is different about the newly unveiled Rembrandt painting?
- How long did it take to create?
Critical Thinking Challenge
Do you think using technology to forge the art of famous painters is a...
Vocabulary in Context
“The data was then used to replicate the artist’s style and generate the facial features for the “Next Rembrandt.””
In the above sentence, the word ...