2-B-20 2種の運動刺激間における知覚的ずれ(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

ThekoaneseJournalofi]sychenomieScience

2003,VDI.22,Ne.1,i)1-52

Prizewinner's

Summary2-B-20

An

illusory

spatial

offset

between

luminance-

and

motion-defined

motioni)

Kazushi

MARuyA*

and

Takao

SATo

Uitiversity

of

Tokyo*

Motion- and

luminance-detined

motion

<MDM

and LDM) were _perceived tobemisaligned when

they were presented physically with the same speed and in_phasc

(perceptual

offset),

Motion-definedmotion isthemotion of patterns

defined

by

the

directien

of the

local

rnotion.

In

thisstudy

we examined the effects of the _global motion speed

(phy$ical

and _perceptual)of MDM and LDM on

theamount of the

_perceptual

offset inorder toexplore the origin of this_phenomenon. The null

point

for

the apparent offset was calculatcd

by

using a _probit analysis,

The

calculated effset

increasedas

the

_{physicalglobal}motion speed

increased.

The

offset phenomenon reported herecan

beinterpretedas aspatjo-temporal conversion of

differences

in

_processing time forLDM and

MDM.

Key werds: vision, motion _perception,second-order motion,

illusory

offset

Several other studies have reported _perceptual

er-rors of the _positionof moving objects

(e.g.

_the

flash

lag

effect,

Nijhawan,

1994).

In

thepresent study, we

report a new Musion thatinduces aperceptual offset

between

two

types

of motion stimulus,

In

this

Mu-sion, _gratings

defined

by

different

attributes were

perceived to

be

misaligned when they were

pre-sented physically with thesame speed and inphase

(Figure

la).We measured the amount of this offset

to explore the origin of _{the phenomenon.}

Spe-cifically, we estimated the null point

between

the

perceptual

and _physica] offset by systematically

varying the _physical offset

between

a

lurninance-definedmotion

(LDM)

and a motion-defined motion

(MDM,

cf.,

Zanker,

l993).

In

addition, we also

in-vestigated

how

the estimated amount of _perceptual

offset was altered while varying the _global motion

speed

(the

speed of moving patterns).

Method

Stimulus The stimuli were

two

types

of square

wave _patterns,

Each

of them consisted of 1024

ran-dom dots ina stirnulus

field

of 120 ×300 _pixels,

t)

_This

_study

_is

_supported

_by

_HFSP

_and

aid ferscientific research from the Ministry of

Education, Science, Sports and Culture for TS,

KM issupported by

JSPS.

* _{Department of}

_Psychology,

_Faculty

_of

_Letters,

University of Tokyo,

7-3-1

Hongo,

Bunkyo-ku,

Tokyo

113--O033

which corresponds toavisual angle of 4x 1Odegrees.

A

square wave _pattern was _generated

by

modu]ating

the directionof the motion of _thc

dots

(up

or

down;

motion-defined rnotion) or their luminance

(lumi-nance-defined motion>, The spatial

frequency

of the

modulation was O.2cfd forboth the MDM and LDM.

For the LDM stimuli the dots were static. For the

MDM

stimuli the

dots

were shifted up or

down

by

8

minutcs of a degree every 20ms. The MDM and

LDM were _generated by shifting thesesquare waves.

The MDM and LDM _patterns were rnoved at the

same $peed, and thephase misalignment

between

the

two stimuli was measured by u$ing the method of

censtant stimuli. The _phase offset was varied inten

steps between a 16 and

-40

deg

_phase angle.

The

pesitiveoffsets

indicate

thattheLDM

is

ahead of the

MDM

in

the direction of the _global motion. The

speed of the _globalmotion was varied infivesteps;

Le,the _pattern was shifted once every

120

ms

by

a

phase angle of

60,

45,

30,

23,

or

15degrees,

The

MDM

stimuli were presented 2.5deg above afixation

point,and the LDM stimuli were _presented 2.5deg

below a fixationpoint,

The

task

of thesubjects was

to

discriminate

the offset

direction

of the LDM by

using a 2-AFC rnethod. The experiment was

con-ducted

in

sessions with a fixed_globalmotion speed.

Each block had 200 trials,and within a

block,

ten

offset conditions were presented

20

times

in

a

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

52

The

Japanese

Journal

of Psychonomic Science VoL 22,No. 1

a

b

:...-.'.::: ...-...

i

Physical

configuration

A o

-en24

_[ tu ¢ m .re

18

a

eM'i2

l･

g6

'8o

8

Figure

1.

(b)

a

De D

eA

AA

o

x

: : :

'

:

.

:

.

:

lPerception

D

eA

e

KMA

YND

vr

2

4

6

8

Globalmotion speed

(deg

visual anglelsee)

{a>A

concept of perceptual offset,

The

amounts of _perceptual offset as a function of thespeed of _globalmotion.

forevery shift condition.

Results

and

Discussion

We

calculated that,as a

function

of

the

physical offset in each conditien, the response rate

for

the

LDM was _greaterthan thatfortheMDM, As aresult,

the response rates of allof the subjects were

larger

than 90% when therewas no _physicaloffset

between

the

LDM

and MDM. This

indicates

that a robust

perceptual offset was experienced by the subjects.

That

is,

the

LDM

was _perceived to

be

misaligned

with the

MDM

in

the

direction

of theglobalmotion,

The null _pointfortheapparent offset was calculated

for

these

data

with a_probitanalysis.

The

amount of

perceptual offset

frorn

the

data

ofthreesubjects was

p]otted as a function of thespeed of theg]obal

rno-tien,

as shown inFigure lb. The shapes of these

curves were very similar

for

allof thesesubjects: the

apparent offset increasedas the_globalmotion speed

increased,

These

results

indicate

that

the

perceptual offset isafunction of the globa] motion speed.

That

is,

the

origin of _perceptual effset

is

probably

the

difference

between

the apparent speeds of theMDM

and

LDM,

or the

difference

intheprocessing time of the MDM ancl

LDM

(cf.

Moutoussis

&

Zeki,

1997;

Nishida &

Johnston,

2002), The speed of a LDM is

supposed

to

be fasterthan that of a higher-order motion stimulus

(Cavanagh

&

Anstis,

2002).

There-forean explanation with such a differenceof appar-ent speed seems tobenatural, However, we believe this

is

not the case.

It

has

been

indicated

by

the

results

from

other experiments,

in

whjch thespeed of

the LDM and

MDM

was matched, thatthe

MDM

was

perceived tobe about 10% fasterthan an LDM with

the same _parameters as the present experiment. If

the origin of the perceptual offset isthe difference

between the apparent speeds of an LDM and an

MDM,

the

LDM

should

be

_perceived

to

be

misaligned with the

MDM

in

theopposite

direction

totheglobal motion. This contradicts the results of the present

study. That is,the origin of the _pcrccptual offsetis

probab]y a

difference

in

the

processing time of the

LDM

and the

MDM.

This

suggests a compatibility

between time and space insome early stages of

vi-sual precessing,

References

Cavanagh, P.&

Anstis,

S.2002 The boogie-woogie

illusion.Rircaption,31,1005-1011,

Moutoussis, K. & ZekL S.1997 A direct

tionof _perceptual asynchrony invision. Pf'oc.R,

Soc.

Lond,

B,

Biol,

Sci,,

264,

393-399.

Nijhawan,

R.

1994

Motion

extrapolation

in

ing.IVdture,370, 256-257.

Nishida,

S,,

&

Johnston,

A,

2002

Marker

dence not _processing latency determines temporal

binding of visual attributes. CurTentBiologlv,12,

359-368.

Zanker,

J.M.

1993

Theta

motioni

A

paradoxical

stimulus toexplore

higher

order motion