Modelling Individual Aesthetic Preferences of 3D Sculptures

Edward Easton, Ulysses Bernardet, Anikó Ekárt

Abstract: Aesthetic preference is a complex puzzle with many subjective aspects. This subjectivity makes it incredibly difficult to computationally model aesthetic preference for an individual. Despite this complexity, individual aesthetic preference is an important part of life, impacting a multitude of aspects, including romantic and platonic relationships, decoration, product choices and artwork. Models of aesthetic preference form the basis of automated and semi-automated Evo-Art systems. These range from looking at individual aspects to more complex models considering multiple, different criteria. Effectively modelling aesthetic preference greatly increases the potential impact of these systems. This paper presents a flexible computational model of aesthetic preference, primarily focusing on generating 3D sculptures. Through demonstrating the model using several examples, it is shown that the model is flexible enough to identify and respond to individual aesthetic preferences, handling the subjectivity at the root of aesthetic preference and providing a good base for further extension to strengthen the ability of the system to model individual aesthetic preference.