Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.750414
Title: Computer graphic control over human face and head appearance : genetic optimisation of perceptual characteristics
Author: Rowland, Duncan Andrew
Awarding Body: University of St Andrews
Current Institution: University of St Andrews
Date of Award: 1998
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The aims of this thesis are two-fold. The first is to develop computer graphics that allow quantitative manipulation of complex visual stimuli. The second is to show that such techniques have utility in the domain of perceptual psychology. There are three main sections to this thesis. The first section creates methods for performing transformations of facial appearance along particular perceptual dimensions. This work begins with 2-D image manipulations and then extends the general principles to 3-D. Effectiveness of the techniques is illustrated with plates showing transformation in age, gender and identity. The second section uses Genetic Algorithms to control the appearance of 3-D computer graphics objects and investigates methods of evolving objects that embody various consumer concepts. Computer graphic models of shampoo bottles are successfully evolved to satisfy a selection of aesthetic and perceptual characteristics. The final section returns to facial stimuli and extends the Genetic Algorithm approach to investigate aesthetic preference for 3-D facial surfaces. The study shows that individual human subjects can evolve facial surfaces based upon their own attractiveness preferences. The faces evolved are non-average and there is consistency between subjects about preferred characteristics. The three parts of this thesis have different theoretical backgrounds and literature relevant to each topic is therefore reviewed at the start of each section.
Supervisor: Perrett, David Ian Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.750414  DOI: Not available
Keywords: QP492.R7 ; Form, solidity, size
Share: