Research Online

Logo

Goldsmiths - University of London

Understanding Aesthetic Evaluation using Deep Learning

McCormack, Jon and Lomas, Andy. 2020. 'Understanding Aesthetic Evaluation using Deep Learning'. In: evoMUSART. Seville, Spain 15-17 April 2020. [Conference or Workshop Item] (In Press)

No full text available
[img] Text (Understanding Aesthetic Evaluation using Deep Learning)
UnderstandingAestheticEvaluationUsingDeepLearning_210120.pdf - Accepted Version
Permissions: Administrator Access Only until 15 April 2021.
Available under License Creative Commons Attribution Non-commercial.

Download (4MB)

Abstract or Description

A bottleneck in any evolutionary art system is aesthetic evaluation. Many different methods have been proposed to automate the evaluation of aesthetics, including measures of symmetry, coherence, complexity, contrast and grouping. The interactive genetic algorithm (IGA) relies on human-in-the-loop, subjective evaluation of aesthetics, but limits possibilities for large search due to user fatigue and small population sizes. In this paper we look at how recent advances in deep learning can assist in automating personal aesthetic judgement. Using a leading artist's computer art dataset, we use dimensionality reduction methods to visualise both genotype and phenotype space in order to support the exploration of new territory in any generative system. Convolutional Neural Networks trained on the user's prior aesthetic evaluations are used to suggest new possibilities similar or between known high quality genotype-phenotype mappings.

Item Type:

Conference or Workshop Item (Paper)

Keywords:

Evolutionary Art, Aesthetics, Aesthetic Measure, Convolutional Neural Networks, Dimension Reduction, Morphogenesis

Related URLs:

Departments, Centres and Research Units:

Computing

Dates:

DateEvent
9 January 2020Accepted

Event Location:

Seville, Spain

Date range:

15-17 April 2020

Item ID:

28095

Date Deposited:

21 Jan 2020 11:40

Last Modified:

22 Jan 2020 17:28

URI:

http://research.gold.ac.uk/id/eprint/28095

View statistics for this item...

Edit Record Edit Record (login required)