Modelling the underlying principles of human aesthetic preference in evolutionary art

Abstract

Our understanding of creativity is limited, yet there is substantial research trying to mimic human creativity in artificial systems and in particular to produce systems that automatically evolve art appreciated by humans. We propose here to study human visual preference through observation of nearly 500 user sessions with a simple evolutionary art system. The progress of a set of aesthetic measures throughout each interactive user session is monitored and subsequently mimicked by automatic evolution in an attempt to produce an image to the liking of the human user.

Semantic aware methods for evolutionary art

“…The goals can be as varied as satisfying the user, as is the case of interactive evolutionary art, satisfy a hard-wired or learned aesthetic criteria [1,5,12,4,11], create novel images [9,3], respond to the "infections" of parasites in a co-evolutionary scheme [6], etc.…”

Feature Selection and Novelty in Computational Aesthetics

“…There are several notable examples of AJSs relying on hardwired aesthetic measures (e.g. [21,12,6]) and also recent works comparing the merits of such aesthetic measures [5,8,1,18]. Usually the AJSs that employ ML, extract information from the images and employ ML to perform aesthetic-based classification or evaluation, which is a common approach in the field of content based image retrieval (e.g., [4,9,17]).…”

Authorship and Aesthetics Experiments: Comparison of Results between Human and Computational Systems

“…There are a great number of papers presenting experiments related to these types of systems. We should highlight the special issue of the Journal of Mathematics and the Arts [4], and some previous papers at this conference, as well as in other fields [1,5,7]. This system would possess a great application by itself when integrated in search engines, as a pedagogical application or as support to artistic researchers all over the world.…”