From Autonomous University of Barcelona –> To University of Bremen
Different methods for classifying art styles in painting have been proposed in artificial intelligence. These approaches provide high categorisation accuracies, but their main drawback is that their explanations to users regarding why a painting is classified in an art style are very limited. To solve this handicap, we have developped the art style painting classifier SHE (Style painting classifier based on Horn clauses and Explanations). This is the first research work that mixes qualitative descriptors and fuzzy logics for art style classification, and provides human-understandable explanations of decisions and outliers. SHE classifies paintings from Baroque, Impressionist and Post-Impressionist styles. We intend to extend in the future this study to include other art styles.
Vicente is a computer Science PhD student (research line: artificial intelligence) and Predoctoral Fellow at the department of Philosophy at the Autonomous University of Barcelona. He also developed part of his research at the Reasoning and Logic Department of Artificial Intelligence Research Institute of the Spanish National Research Council. He is also the COO and editor of Matesfacil, a web for learning maths until high school level. His interests range from the foundations of mathematical fuzzy logic to artificial intelligence and cognitive science. In this way, his main theoretical research has been focused on the universal Horn fragment of predicate fuzzy logics, while his current principal activity is the study of cognitive qualitative fuzzy descriptions of colours and its applications.