Can a computer distinguish Kandinsky’s “To the Unknown Voice” from Van Gogh’s “Sunflowers”?
European researchers are working on it. The researchers analyzed various artificial vision algorithms used to classify art, and found that certain aesthetic measurements—such as analyzing pixels and color distribution and diversity—can help the computers determine artistic style, said Miquel Feixas, one of the authors of a report of the team’s findings. According to the researchers, software powered by certain mathematical algorithms allow a computer to differentiate between artistic styles based on low-level pictorial information. Low-level pictorial information includes brush thickness, material type, and palette. “It will never be possible to precisely determine mathematically an artistic period nor to measure the human response to a work of art, but we can look for trends,” Feixas said. But, proving a computer is no discerning art critic, the algorithms still aren’t advanced enough to classify art using medium and high-level concepts. People, on the other hand, can comprehend conce
