MAPPING EMOTIONS TO SHAPE PARAMETERS USING ARTIFICIAL NEURAL NETWORK

glassesbeepingΤεχνίτη Νοημοσύνη και Ρομποτική

20 Οκτ 2013 (πριν από 3 χρόνια και 9 μήνες)

83 εμφανίσεις

MAPPING
EMOTIONS
TO

SHAPE PARAMETER
S

USING
ARTIFICIAL NEURAL NETWORK


Khusnun Widiyati
1
, Hideki Aoyama
2

Department of System Design Engineering, Faculty of Science and Technology

3
-
14
-
1 Hiyoshi, Kohoku
-
ku, Yokohama, 223
-
8522 Japan

1
Email:
khusnun@ina.sd.keio.ac.jp

2
Email:
haoyama@sd.keio.ac.jp



Abstract.

Aesthetic and emotion are becoming an attractive factor for product design.
T
he product image in customer

s mind is a result from form impression process that
incorporates

both aesthetic and emotion during product design.

And therefore,
there

is a
need for sophisticated design

based on terms expressing

Kanse

, which in this study
these terms are defined as

Kansei words

.
Though researchers are trying to formalize
the rela
tionship between shape information and the emotions, the knowledge about the
relation between shape parameters to the associated evoke emotion
is

still limited.
T
his
paper

attempts to

propose an
approach

to map emotion
s

to
product
shape

parameters

using ar
tificial neural network.
To achieve this, recognition on product shape
parameters
and collection of emotion/ Kansei words were becoming the first important
step. S
everal 2D models were
then
created

based on the shape parameters
.

Assessment
of product image
s was
carried out by presenting the model and emotion/Kansei words
in
semantic differential survey
to
participants from
design engineering student
s
.

Artificial
neural network was used to map the product image to shape parameters.
A
n evaluation
of how users

perceived the shapes was
conducted

to validate the artificial neural
network model.
A case study on
PET bottle

also presented to give clear figure how
the
mapping method works
.


Keywords: Aesthetics, emotion/Kansei, artificial neural network, shape
parameters