Sep. 28, 2012 at 1:05 PM ET
While art may be in the eye of the beholder, a computer algorithm has been developed that can classify paintings as well as distinguished human art historians.
The task is a breakthrough for the field of artificial intelligence, showing that futuristic computers won’t just be trivia-answering champs such as IBM’s Watson or math nerds such as the Todai Robot, but trusty companions in fashionable art circles as well.
The algorithm computes from each painting 4,027 numbers that reflect the content of the image such as color, texture and shapes and then uses pattern-recognition and statistical methods to analyze similarities and differences between painting styles.
In an experiment, Lior Shamir and Jane Tarakhovsky of Lawrence Technological University used their algorithm to analyze about 1,000 paintings from 34 well-known artists. The computer provided a network of similarities between painters that agrees with the perception of art historians.
For example, it was able to distinguish between classical realism and modern artistic styles and then identified sub groups of painters that were part of the same artistic movements, such as grouping the painters Raphael, Leonardo da Vinci and Michelangelo close to each other.
“These results demonstrate that machine vision and pattern recognition algorithms are able to mimic the complex cognitive task of the human perception of visual art,” the researchers conclude in a recent paper on their work published in the Journal on Computing and Cultural Heritage.
This skill, they add, “can be used to measure and quantify visual similarities between paintings, painters and schools of art.”