Researchers at Rutgers University have developed an artificial intelligence (AI) system that can identify an art forgery by examining the piece's brush strokes.
The new system analyzed nearly 300 line drawings by Picasso, Matisse, Modigliani, and other famous artists and created a dataset of about 80,000 individual strokes.
A deep recurrent neural network then learned what features in the strokes were important to identify the artist.
In addition, the researchers trained a machine-learning algorithm to look for specific features, such as the shape of the line in a stroke. The data yielded two different techniques to detect art forgeries, and the combined method was able to correctly identify artists about 80% of the time.
The researchers commissioned other artists to create drawings in the same style as the pieces in the dataset, and they found the system was able to identify 100% of the forgeries by looking at a single stroke.
From Technology Review
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA