Aesthetics and evaluation of objects is becoming increasingly important in contemporary society. Although there have been many studies on processes related to computational aesthetic, a clear formalisation and visualization of the aesthetic field is still lacking. In this paper, we present a set of Machine Learning techniques and mathematical methods to extract the most important features related to aesthetical evaluation, thus making this process automatic, without the human intervention. The techniques are then applied to a sample of 83 images of triangles, produced by artists. The results of the empirical method provide a series of measurements that allow the extrapolation of mathematical aesthetic characteristics of the images and their location in the aesthetic space.
Shaping the aesthetical landscape by using image statistics measures
Bertacchini F.;Pantano P. S.;Bilotta E.
2022-01-01
Abstract
Aesthetics and evaluation of objects is becoming increasingly important in contemporary society. Although there have been many studies on processes related to computational aesthetic, a clear formalisation and visualization of the aesthetic field is still lacking. In this paper, we present a set of Machine Learning techniques and mathematical methods to extract the most important features related to aesthetical evaluation, thus making this process automatic, without the human intervention. The techniques are then applied to a sample of 83 images of triangles, produced by artists. The results of the empirical method provide a series of measurements that allow the extrapolation of mathematical aesthetic characteristics of the images and their location in the aesthetic space.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
