Software that Authenticates Art

(Thanks AJ). A new set of software tools, developed by a Dartmouth College team seems to be able to "fingerprint" the style of a painter's work and thus be able to detect fakes and imitations.

"There are properties in an artist's pen and brush strokes that aren't visible to the human eye, but that are there nonetheless. And we can find them, through mathematical, statistical analysis," said Dartmouth computer science professor Hany Farid, who developed the algorithms, along with math professor Daniel Rockmore and graduate student Siwei Lyu."

The article warns that:

"Museum curators and statisticians caution that the Dartmouth group's techniques have only begun to be tested. Using algorithms to back up scholars' suspicions is one thing; uncovering a fraud with just a computer, that's completely different. And in the art world, no scientific method is considered as sure as the eye of a seasoned connoisseur.

"This is very unusual," said Nadine Orenstein, the curator of the drawings and prints department at the Metropolitan Museum of Art in New York. "We're all a bit skeptical."

Other artworld comments were just as skeptical:

"I’m highly skeptical of using the computer for this kind of approach," said Laurence Kanter, curator in charge of the Robert Lehman collection at the Metropolitan Museum of Art in New York City. "It’s very possible that a program like that could be exploited to great advantage, but I still need to be persuaded."

I suspect that they're all a bit scared actually, because they're about to (a) lose the gig of having the power to be able to authenticate something just by being a "seasoned connoisseur" and (b) about to (possibly) have an avalanche of their past "authenticated works" be reported as fakes.