Artificial intelligence in Italian museums will judge the popularity of artists by the facial expressions of visitors

Researchers hope this technology will help organize exhibitions better, but creatives fear it will fuel a race for likes. If you smiled at Picasso in a museum but frowned at Pollock, new technology will soon be able to track your facial expressions to judge the artist’s popularity.

The researchers presented tests of artificial intelligence camera systems in three Italian museums, aimed at assessing the response to individual works of art.

A technology that can read five facial expressions - happy, sad, neutral, surprised or angry - has been tested for three weeks in Rome, Bologna and Parma. It can also detect a person’s gender, age, and eye movement.

Researchers hope that curators can use this information to change the design of their exhibits, promoting works they like, potentially discarding ones of less interest.

However, the algorithm was met with a mixed reaction: some artists and curators expressed concerns that this could hinder viewers or even turn art into another race for "likes".

On the first day of the system’s operation, there was no indication at the Bologna City Museum that the technology was working, other than small black cameras attached to the walls and a disclaimer at the box office.

Inside the exhibition hall, the visitor came to a painting depicting Saint Sebastian of the 12th century, who was looked after by Saint Irene. The eye height sensor registered a slight smile and sent a "happy" report back to the database. It was recorded that the visitor briefly glanced at Irene before looking directly at Saint Sebastian.

“I was surprised what a great opportunity it was,” curator of the Bologna City Museum, Silvia Battistini, told the Telegraph. "As a curator, you usually don’t get this much up-to-date information about the public’s point of view without some kind of intermediary."

Museums in Washington and London have previously invited visitors to interact with artificial intelligence at their exhibitions, but ShareArt’s experiment is the first that doesn’t require active participation from art lovers.

A research team from Italy’s National Agency for New Technologies, Energy and Sustainable Economic Development initially set up a system to help the country reopen museums while avoiding potential Covid 19 contamination. Cameras could alert staff if visitors stood too close to each other or removed their masks….

According to Ricardo Scipinotti, the engineer who helped create it, the facial recognition system still sees "a lot of neutral" facial expressions. Visitors do not need to worry about privacy as the images are not saved and the data is displayed “as numbers for analysis”.

Priya Khanchandani, head of curatorship at the Design Museum in London, says that while residence time sensors can help art institutions understand their audience, she will not use facial recognition. “Facial expressions are a natural science, not an exact science,” she told the Telegraph. “It would be a shame if data recording our intimate ways of reacting to a work of art limited the spontaneity of our behavior. The idea that our emotions can be ’stored’ is frustrating and a little confusing. "

Contemporary artists Rob and Nick Carter embrace this technology. “It may sound intrusive, but this information is invaluable to artists like us as it triggers dialogue and audience exploration that could potentially initiate entirely new work,” they said.