Computer Can Tell Jackson Pollock Paintings From Pretenders

February 24th, 2015

Recognizing and interpreting art is a hot topic in artificial intelligence research, and of special importance to the detection of falsified artworks. A researcher at Lawrence has created a set of computer algorithms that can differentiate between Jackson Pollock’s chaotic paintings and those produced by other artists imitating his style.

To untrained eyes, the paintings of Jackson Pollock’s seem nothing more than randomly splattered paint on canvas. But Pollock’s style is more than merely random: it is intimately expressive of the  unique proportions of his body and mind. But Lior Shamir, a computer scientist at Lawrence Technological University has been able to teach a computer to make the distinction:

For the Pollock paintings, Shamir obtained 26 paintings known to be by the artist and a second set that was painted by artists who were inspired by Pollock and attempted to create works in his style. The pieces were normalized to contain 640,000 pixels and then divided up into 16 equal-sized areas. Twenty works were used to train the software, then the remaining six were used for testing purposes. The analysis was then repeated multiple times, each with a different set of 20 training images, in order to provide a greater statistical power. The piece was determined to be Pollock or not based on a majority voting system, with each of the 16 sections of the painting getting one “vote.”

Read the full article here on Ars Technica.