Modelling creative decision making

Investigate aesthetic pleasantness in the visual domain, in an inter-disciplinary manner.

Disciplines:

  • Cognitive Science
  • Computer Vision
  • Visual Arts

The project draws inspiration from neuroscience and psychophysics to develop an artificial intelligent system that learns aesthetic preferences of a population. Using datasets of photographs rated by humans depending on aesthetics, the artificial intelligent system extracts brain-inspired features from those images to determine what is considered as aesthetically good and aesthetically poor. The eventual applications for this artificial intelligent agent are many, with for example, the possiblity of filtering images for an online community. Moreover, it is hoped that the brain-inspired features used can help to learn more than simple photography rules and actually embed more general aesthetics rules, allowing to explore the agent on other types of visual medias. 

The goal of the project is to improve the understanding of what is aesthetically pleasant to the human visual system, define key features that would allow to evaluate beauty and establish a strategy to create attractive and novel visual shapes.

Research Fellow
Francois Lemarchand
Supervisors

Roman Borisyuk, Haline Schendan, Giorgio Ganis (Plymouth University)