2-A-12 三次元物体認識における空間的処理容量の個人差と視点依存症(2002年度 日本基礎心理学会第21回大会優秀発表賞)

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

ThelaPanesejburnalofPsychonomicScience

2003,VoL 22,No.1,43-44

Prizewinner's

Summary2-A-12

Individual

differences

of

spatial

span

and

view-dependence

in

three-dimensional

object

recognition

Takeo

KoNDou

and

Tamotsu

TosHiMA

Hiroshima

Uheiversity*

A largememory set size and a high similarity have been reported tocause view-dependence

in

3-D object recognitien

(NewelL

1998), Inthisstudy, we examined whether individual

differences

in the capacity of spatial information processing

(spatial

span) relate

to

this

type of

view-dependence

or not.

According

to

the

scoring of a spatial span task

(Shah

& Miyake, 1996),ten

participantswith a

high

spatial span score and ten_participantswith a low spatial span score were

assigned tohigh and low spatial span groups,respectively,

The

results indicatedthatonly the

low

spatial span _group showed a view-dependence

in

thc

high

cognitive

load

condition of the 3-D objcct

discrimination

task. This suggested thatindividual

differences

in

spatial span could beone of the

factors

which affect the view-dependence in3-D object recognition,

Key words: view dependefice, canonical view, three-dimensional object recognition, working memory

Introduction

Working

mernory

<WM)

isa limitedresourcc that

processes actively stored

information

to _perform

goal-orientedtasks,and thereare individualand

de-velepmental differences.Simultaneously

increasing

in

thenumber of target objects

(memory

set size) and thesimilarity

between

targetand

distracter

objects,

makes therecognition of 3-D objects view-dependent

CNewelL

1998),

This

means that the

increased

load on _processing

and storing the 3-D objects _produces a

view-dependent _performance. Individual

differences

also

eccur invisual-spatial WM, and a span testtoassess the capacity of an individuaVs WM has been

devel-oped

(Spatial

span task;

Shah

&

Miyake,

1996).

The presentstudy examined whether the

perform-ance of

3-D

object recognition would change with

individual

differences

in

WM

capacity.

Method

Participants Thirty-two naive undergraduate

and graduate students were volunteered to partici-pate.*

Department

of

Psychology,

Hiroshima

sity

Kagamiyarna

1-1-1,

Higashi-Hiroshima

739-8524

Materials The stimuli were twenty objects

from

a set,some of which are shown inFigure 1(a),The

set of target

images

consisted of a canonical view

which was

decided

by

the participants and another two views which were incrernented from the canoni-cal view

by

40'

and

80e

respectively, Thc

distracter

objects had a main

body

of the same shape as the

target

objects, and had appendages at approximately

the same locationand with the same

interrelation-ship,

The

set of distracterimages were made from

the distracterobjects and

from

thcsame viewpoint

as the targetimage set.

ExperimentalDesign&Procedure

Thedesignof

the present study used

the

spatial span task

(SST)

score _group as a

between-participants

variable

(HIGH,

LOW>, and the memory set size

(LARGE,

SMALL)

and viewpoint

(OO,

400,800)as the within-participantvariab]es,

First,a

SST

was used toassess theability of the

participant'sworking memory capacity to

simulta-neously process and store spatial infermation,

In the

following

delayecldiscrimination task,the

participants were

initially

asked to rernember the

shape of

targets

which were rotated

five

timeson

the

display

of a

CRT.

According tothememory set size condition, theparticipants were shown threeor one

targetobject.

The

_participantsthen were shown all

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

44 The

Japanese

Journal

of Psychonomic Science VoL 22,No. 1

a

Figure

1.

(a)

Examples

of the stimuli,

(b)

Med

stimulus object's viewpoint, with

SST

score

LARGE>

as a _parameter,

b

I:9vN

IZOO 1100 1000 900 800

ian reaction times as

(LOW,

HIGH)

group

the

views

from

thirty-six

viewpoints of the target

object, and asked tochoose the best

_(canonical

view:

O")

frem

these views. After･

that,

the _participants were asked to

discriminate

the target

images

from

the distracter

images

by

pressing a computer mouse

button

(regardless

of the changes in viewpoint).

Each trialwas composed of a fixationcross

for

500

ms,

then

a

display

of a test

image

(the

target or

distracter),

The testimage disappeared when the

participantresponded, and thenext tria]

_(which

was

followed

by

blank

for

500

ms>

then

started,

Results

Ten

participants with a high

SST

score and ten with a low

SST

score were assigned totheHIGH and

LOW

score groups, respectively. The remainder of

theparticipantswere excluded frornour analysis.

The median reaction

times

for

the correct

re-sponses as a

function

of viewpoint, with the SST

score group and memory set size as a _parameter, are

shown

in

Figure 1(b}.

A three-way ana]ysis of variance was calculated

with one

between-participants

factor

(SST

_group;

HIGH,

LOW)

and two within-participant factors

(memory

set size: LARGE,

SMALL:

viewpoint: OO,

400,

80e}.

The

analysis revealed a significant

interac-tionof al]of the threefactors.Ina second analysis,

only

for

theLARGE memory set size condition, the

LOW _group'sreaction time

for

800

was significantly

longer

than

for

O".

Further analysis revealed that

therewas not a significant effect of viewpoint for

the

O 40 80

VIEW

(degl

a

function

of the angles of

the

and memory set $ize

{SMALL,

SMALL

memory set sizeand the HIGH group.

Discussion

The results revealed that only the LOW scere

group showed a view-dependence

in

the condition

with a LARGE memory set size and close similarity

between

the objects, thatis,the condition of a

high

load on the

WM

The

view-dependence observed in

the current study was considered to be caused by this

increased

]oad on theWM. The _performance of

the participants thereiorechanged with their indi-vidual

differences

in

the capacity of theirWM

indi-cated

by

their

SST

score. Itisinterestingto note that,aside frornthe factorsof an object's attribution,

internal

factors

of an observer can cause the

occur-rence of view-dependent recognition, Individual

dif-ferences

in

WM could thereforebe important factors

when examining 3-D recognition _processes,

because

a _{participant'sperformance} of recognizing the same

object might vary with theirindividualdifference

in

WM.

References

Newell,

F.

N.

1998

Stimulus

context and view

pendence inobject recognition. Il2rception,27,

68.Shah,

P.

&

Miyake,

A.

1996

The

separability of

working memory resources

for

spatial thinking

and

language

processing: An irrdividual ences approach.

Iburnal

of

Expen'mental