In the museums of Rome, Bologna and Parma, the ShareArt computer vision system is being tested, which evaluates the reaction of visitors to individual works of art. The model classifies emotions based on five facial expressions-happy, sad, neutral, surprised or angry, and also determines the gender, age and eye movement of visitors.
The researchers believe that curators will be able to use this information to improve the visitor experience of exhibitions, drawing attention to the most popular art objects from them and removing those that cause less interest. Previously, museums in Washington and London invited visitors to interact with artificial intelligence in their exhibitions, but the ShareArt experiment is the first that does not require active participation of visitors and works completely independently.
In addition to the functionality described above, the cameras warn the museum staff if visitors are too close to each other or remove their masks. The data received from ShareArt cameras is confidential. In particular, they do not perform face recognition, and the photos of visitors are immediately deleted after processing by the neural network. The artists appreciated the new technology, as it provides a valuable tool for analyzing the impact of art on people and brings the dialogue between people and art to a new level.