According to the researchers, artificial vision algorithms make it possible to differentiate artistic styles based on low-level pictorial information , which includes aspects such as the thickness of the brush, the type of material and the composition of the color palette. Human categorization strategies, however, also include mid-level concepts, to differentiate the specific objects and scenes that appear in a painting and the type of painting (landscape, portrait, still life …), and high-level concepts, which they take into account the historical context, as well as the knowledge of artists and artistic trends.
Beyond the implications for philosophy and art, scientists want to apply their research to the development of image visualization and analysis tools, the cataloging of collections in museums, the creation of public informational and entertainment facilities, as well as to a better understanding of the interaction between people, computers and works of art .
