computer coded mona lisa

s17_Mona-Lisa-Final_inline.jpg

Illustration by Peter and Maria Hoey

A fun aspect of appreciating a painting is understanding the influences behind it. That’s the business of art historians, who soon may be doing their sleuth work faster. At the Art & Artificial Intelligence Lab at Rutgers University–New Brunswick, professor Ahmed Elgammal and Babak Saleh GSNB’13,’17 have created an algorithm that allows a computer to evaluate paintings like an art scholar—but at the speed of light, identifying and comparing thousands of paintings for considerations like style, color, and composition, leading to discoveries such as the influence of French impressionist Frédéric Bazille on Norman Rockwell.

VIEW A VIDEO, PRODUCED BY THE STATE OF THE ARTS NEW JERSEY.