A QUANTITATIVE MODEL FOR IDENTIFYING REGIONS OF DESIGN VISUAL ATTRACTION AND APPLICATION TO AUTOMOBILE STYLING
DS 84: Proceedings of the DESIGN 2016 14th International Design Conference
Year: 2016
Editor: Marjanovic Dorian, Storga Mario, Pavkovic Neven, Bojcetic Nenad, Skec Stanko
Author: Pan, Y.; Burnap, A.; Liu, Y.; Lee, H.; Gonzalez, R.; Papalambros, P. Y.
Series: DESIGN
Section: INDUSTRIAL DESIGN
Page(s): 2157-2174
Abstract
Analysis of design regions of visual attention that affect aesthetic appeal is an important topic for both practicing designers and design researchers. The paper introduces a data-driven methodology consisting of four stages: (1) design feature learning, (2) design attribute prediction, (3) salient feature selection, and (4) salient feature visualization. Using this methodology, we making inroads to inverting the nonlinear function from design images and design aesthetic attributes, and give preliminary results for an automotive styling study.
Keywords: visual attention, design aesthetic attribute, quantitative model, data-driven, automobile styling