Assessing Aesthetics of Generated Abstract Images Using Correlation Structure

Can we generate abstract aesthetic images without bias from natural or human selected image corpi? Are aesthetic images singled out in their correlation functions? In this paper we give answers to these and more questions. We generate images using compositional pattern-producing networks with random...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE Symposium Series on Computational Intelligence (SSCI) pp. 306 - 313
Main Authors Khajehabdollahi, Sina, Martius, Georg, Levina, Anna
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2019
Subjects
aesthetics
Biological neural networks
Computer architecture
Correlation
correlations
CPPN
Image color analysis
image generation
image statistics
Network architecture
neural network
Neurons
Shape
Online AccessGet full text
DOI10.1109/SSCI44817.2019.9002779

Cover

More Information
Summary:Can we generate abstract aesthetic images without bias from natural or human selected image corpi? Are aesthetic images singled out in their correlation functions? In this paper we give answers to these and more questions. We generate images using compositional pattern-producing networks with random weights and varying architecture. We demonstrate that even with the randomly selected weights the correlation functions remain largely determined by the network architecture. In a controlled experiment, human subjects picked aesthetic images out of a large dataset of all generated images. Statistical analysis reveals that the correlation function is indeed different for aesthetic images.
DOI:10.1109/SSCI44817.2019.9002779