A method of kansei acquisition to 3D shape design and its application

Abstract

For personal design of HUTOP (a human-ergonomics-oriented design method), a support system like a "design adviser" which customizes the design with customer's kansei is indispensable. To realize this system, the technology that acquires customer's taste and changes customer's request into the physical parameter for the product design is needed. Although some methods for such system have been proposed, the methods that consider customer's individuality and also handle 3D shape are few. Because of this point, we propose a method of kansei acquisition to 3D shape. The method is based on 3D shape deformation and machine learning techniques, and can model the relationship between 3D shape and kansei about them. Sofa was chosen as an example of design and an experiment for testing the method was made, and then some problems were made clear.