Interactive Reduct Evolutional Computation for Aesthetic Design

Abstract

We propose a method of evolving designs based on the user’s personal preferences. The method works through an interaction between the user and a computer system. The method’s objective is to help the customer to set design parameters via a simple evaluation of displayed samples. An important feature is that the design attributes to which the user pays more attention (favored features) are estimated using reducts in rough set theory and reflected when refining the design. New design candidates are generated by the user’s evaluation of design samples generated at random. The values of attributes estimated as favored features are fixed in the refined samples, while other attributes are generated at random. This interaction continues until the samples converge to a satisfactory design. In this manner, the design process efficiently evaluates personal and subjective preferences. The method is applied to design a 3D cylinder model such as a cup or vase. The method is then compared with an Interactive GA.