バイオロジカルモーション知覚中心視と周辺視における検出能力(第23回大会 優秀発表賞抄録)

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

7)beJaPanesefoztrnatofRsychonomicSciencE

2005,VoL 24,No.I,135-136

Summary

ofAwarded

Presentation2P47

Eccentricity

dependency

of

the

-

_perceptlonof

biological

motioni)

HanakoIKEDA,*i

***

_Randolph

_BLAKE,**

_and

_Katsumi

_WATANABE***

Uitiversity

of

Tsukuba.*

Vanderbilt

Universit.v,'*

Ndtionat fnstitute

of

Advanced industriatScienceand 7kechnology***

In

thisstudy we compared the performance of

detecting

biological

motion

in

the

foveal

and pcripheral visual

ficlds.

Correct

and scrambLed

biological

motion were successively presented and

the subjects indicated which of the two intervalscontained the biologicalmotion. Detectlon

performance

(defined

as resistance against noLse) was determined by using astaircase method,

In

the

foveal

and peripheral visual

fields

detection

performance was saturated as

the

stimulus size

was

increased.

However

the performance of

detecting

motion

in

the peripheral

fields

was not

compensated by spatial rnagnification. These results suggest that theresource, or mechanism, for

biologicalmotion _perception isconfined tothecentral rcgion of thevisual field.

Key words: biologicalmotion, eccentricity, visual

Introduction

Inhis original study

_Johansson

₍₁₉₇₃₎

_presented

an

individual,

with small lightsources attached to

themain

joints,

who was walking indarkness so that

allether visual information was removed. But even

though the visual

information

was

highly

impover-ishedobservers could easily recognize a person

walk-Lng.

This

ability has

been

ca]Ied "biological

motion

perception".

Following

Johansson's

study many

oth-ers

have

investigated

the perception of biological

motion.

However

there

has

not

been

any study of

the ability todetectbiologicalmotion inthe periph-eral visual

field.

The

irnportant

_purpose of

biological

motion p¢rception maybe toselect

biologically

rele-vant targets across the entire visual field,The

present study exarnined eccentricity

dependency

of

biological

motion _perception.

1)

A

furl

report of thisfinding has appeared

whcre

(Ikeda

et a],,2005).

*

Graduate

School

of

Comprehensive

Human

ences, Univcrsity of Tsukuba, 1-1-1 Tennoudai,

Tsukuba, Ibaraki305-8574

Japan

** _{Department of Psychology, Vanderbilt}

sity, 511 Wilson HaiL 111 21st. Avenue South,

Nashville,

Tcnnessee

37203,

USA

***

_Visual

_Cognition

_Group,

_Institute

_forHuman

ence and

Biomedical

Engineering,

National

tute of

Advanced

Industrial

Science

and

nology,

AIST,

Tsukuba

Centrai

6,1-1-1

Higashi,

Tsukuba, Ibaraki305-8566

Japan

Materials

and

Methods

Subjects

Six

subjects

₍₂

temale,4 male) _partici

pated

in

the study,

Stimuli Biologicalmotion scquences were made

from videotapes of an

jndividual

_performing

five

activities

(jumping,

running. walking,

kickLng,

and

throwing a ball).The videotapes were digitizedat 25

Hz. The biologicalmovements were represented

by

the movement of 12

black

dots,

which were major

joints,

on a white background and the stimuli were

presented

in

7sizes of visual angle

(O.5,

I,2,

4,

8,

12,

and 16degrees).The stirnuli were also _presented at

3

eecentricities

(O,

4,

and

12

degrees}.

The

stimulus size was changed

by

magnifying allof the spatial dimen-sions of the visual stimulus, thati$,the dot size was

increased.Invertedbiologicalmotionswerealsoused.

Procedure

Subjects

were

instructed

tomaintain

a

firm

fixationduring each triaL A single trial

con-sisted of two 800mMisecond intervalsof stimulus

presentation which were separated bya

blank

period

of

500

milliseconds.

One

stimulus

interval

contained correct biologicalrnotion

(either

upright or inverted).

The other stimulus intervalcontained a scrambled version created by randornizing

the

starting _position of the

dots

of the correct stimulus within the area

of the

biological

motion stimulus. Thc presentation

order of the stimuli was randomized. Afterviewing

two intervalsthe subjects reported which stimu]us

interval

contained the correct

biological

motion

by

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

136

The

Japanese

Journal

of

Psychonomic

Science

V

pressinga keys of

the

computer

keyboard

₍₂

alterna-tives,forced choice).

The

_performance

of detecting

biological

motion was defined as resistance against

motion noise

(Cutting,

Moore,

&

Morrison,

1988), A staircase method was used tomeasure the noise re-sistance:

3

consecutive correct responses added 4

dots

tothe noise level,and 1 incorrectresponse

re-duced thenoise level

by

4

dots.

When

the

number of

reversals reached 18,a session was finished. The

performance

in

a session was determined by

averag-ing

the noise

levels

of the last

5

reversals. _Three subjects were testedwith allcombinations or the 2 stimulus

types

(upright

and

inverted),

7 stimulus sizes,and 3 eccentricities, which resulted in42

condi-tions, For the other3subjectsthelargeststirnulus

size

(16

degrees)

and

3

eccentricjties were used 6

con-ditions.

Each

subject _performed 4 sessions of for

each condition, and the _performances

for

each

con-ditionwere averaged,

Results

and

Discussion

The

results

from

the 3 subjects who tested inall conditiens are _presented inFigure 1(a).

The

perfor-mances

(defined

as noise resistance) were averaged

and plottedas a

function

of the size of thestimulus,

Inallof thetypesef stirnulus conditions the

perform-ance increased as

the

stimulus size

increased

and as

the eccentricity

decreased,

However, the

perform-ance levelssaturated at certain stimulus sizes.

This

was true forall of the 3eccentricities.

Most

impor-tantly,themaximum levelof performance was

diffe-rent

for

different

eccentricities, Inother words, the

stimulus magnification did not compensate

for

the

reduced _perforrnance with peripheralviewing.

This

discrepancy

in

thesaturation levelsfordifferent

ec-centricities was more apparent fortheupright

condi-tion. The _performance of

the

subjects was

higher

forthe upright condition than

for

the inverted

con-dition

(i.e.

inversion

effect;

SumL

1984). However,

thc

inversion

effect appeared to diminish

_for

the

peripheralvjsual field.

Figure

1{b}

illustrates

the performance as a

func-tionofeccentricity, averaged forthe6subjects who

performed the task with a stimulus size of

16

de-grees, The perfermance was

higher

for

smaller

ec-centricities and

for

the upright condition. These

(a)

_!

g

･aj

g

g

= oL 24,No, 1 Uprignt 1

:T:

}gg･

gsocaneiD

[K"O O 2 46 S 1or21461e

[HO

moosuuaswoo10o

SOO24fiBtOtZ4161S

am

tnverted [KY4q O 24 6 SriO1214rl61S

!rugo'8ggSsuee...･e"""

!Oo246

101X416te simulussize(deg)

(b)

45g,,13sE3o{25=i2o t5 stimdussizetdeg) +LVight-O-irNeted 10 e 4 12 eccenticitvideg)

Figure 1,

(a)

The

results of 3 subjects whe

tested in all conditions.

_(b)

The averaged

perforrnance of the task with a stimulus size

of 16 degrees

is

shown. Adapted from Ikeda

et aL

(2005)

with perrnission

from

Elsevier. results suggest that the resource forthe _perception

Qf biologicalmotion is not uniformly

distributed

across the visual

field

but

rather,

is

limited

to the

central region of thevisual field.

References

Cutting,

J.

E,,Moore, C. & Morrison,

R.

1988

ing

the motions of

human

gait. Pereoption &

RsychoPdysics,

44,

339-347.

Ikeda, H. B]ake,

R.

Watanabe, K. 2005 Eccentric

perceptionof bio]ogicalrnotion isunscalably _poor.

Vision

Research,

45,1935-1943.

Johansson,

G,

1973 Visual _perception of

biological

motion and a model for

its

analysis.

PercePtion

&

Rsychopdysics, 14,201-211.

Sumi, S. 1984

Upside-down

presentatien of the

hansson moving ]ight-spot

_pattern.

RercePtion,13,