A neural network became a expert at detecting art forgeries by learning how famous artists drew their line strokes. Researchers at Rutgers University and the Atelier for Restoration and Research of Paintings in the Netherlands used a sample set of 300 line drawings from well-known artists such as Picasso and Matisse. From those drawings, the AI examined 80,000 lines strokes and learned what characteristics in the strokes were unique to the different artists.
From Technology Review:
The researchers also trained a machine-learning algorithm to look for specific features, like the shape of the line in a stroke. This gave them two different techniques to detect forgeries, and the combined method proved powerful. Looking at the output of the machine-learning algorithm also provided some insight into the RNN, which acts as a “black box”—a system whose outputs are difficult for researchers to explain.
Since the machine-learning algorithm was trained on specific features, the difference between it and the RNN probably points to the characteristics the neural network was looking at to detect forgeries. In this case, it was using the changing strength along a stroke—that is, how hard an artist was pushing, based on the weight of the line—to identify the artist. With both algorithms working in tandem, the researchers were able to correctly identify artists around 80 percent of the time.
The researchers also commissioned artists to create drawings in the same style as the pieces in the data set to test the system’s ability to spot fakes. The system was able to identify the forgeries in every instance, simply by looking at a single stroke.
“A human cannot do that,” says Ahmed Elgammal, a professor at Rutgers and one of the paper’s authors.
Image: Picasso, Matisse, or a Fake? Automated Analysis of Drawings at the Stroke Levelfor Attribution and Authentication
Editor’s Note: Richard Metzger is a connoisseur of cannabis, and recently started growing his own. He’s test-driving high-end rig good for small-scale grows from Cloudponics. This is not a sponsored post, Boing Boing is not getting anything from Cloudponics. Metzger’s just really *that* enthusiastic about weed, and so far he likes the Cloudponics setup. Here’s […]
Forbes publishes 300 stories a day, and is developing AI software that writes first drafts of articles. From Digiday: With Bertie, a contributor who writes regularly about the automobile industry might open up the tool to find the makings of an article about Tesla, complete with links to relevant, related articles published both on Forbes […]
Remove.bg, a new free web app released by developer Benjamin Groessing on Monday, keeps popping up in my feeds so I thought I’d try it out. With one click, it removes the background of a photo using AI: Today I'm launching https://t.co/PX30RBnfSm – a tool to automatically remove the background of any image (with a […]
Things move fast in the world of social media, and they don’t slow down for marketers looking to make an impact. Whether you’re grabbing eyeballs for a major company or a local business, you’ll need to adjust your strategy for every platform. Don’t have a strategy to begin with? That’s where the Social Media Marketing […]
It’s a rude awakening for that rookie vacationer abroad when they try to plug in their gear for the night. Veteran jet-setters know that outlet shapes can vary wildly from country to country, which necessitates that most boring must-have for any world-traveler: A sackful of clunky power adapters. Awkward problem, elegant solution: The Twist Plus […]
Looking for a career in music behind the boards, either as a music producer or DJ? It’s a good bet that you’re going to be working with Ableton Live. Each new iteration of this powerful workstation gives the user more tools to create, and it’s just as well suited for the task of meticulous track […]