Enhancement of horizontal motion detection sensitivity by vertical illusory motion component(Summary of Awarded Presentation at the 27th Annual Meeting)

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

1}ze_{pma.neseYbumalof}Rsychonomic Sciance

2009,VoL 28,No.1,183-184

Summary

of

Awarded

Presentation2P59

Enhancement

of

horizontal

'

by

vertical

illusory

motion

detection

sensitivity

-motlon

component

HirornasaTAKEMuRA*

and

Ikuya

The

[iniversit]s

of

7bleyo"MuRAKAMI*

Visual

motion

inforrnation

passes

through

several

distinct

processing stages,

but

the

stage

where the perceptua]

limit

of motion dietecttonarlses

is

stM unknown,

We

tested

how

vertical

illusoryrngtion affects the detection of horizontal motion. We _presented a centra] Gabor _patch moving horizontally,together with asurrounding _grating moving vertical]y.

The

central stimulus

is

perceived as moving obliquely

by

integration

betvveen

thephysical and

illusory

motion

induced

by the surrounding motLon. The _participants were asked to

judge

the

horizontal

mot'ion

compo-nent of thecentral stimulus

Oeft

vs. right). We found motion detection_periormance was enhanced

at a moderate surrounding speed

in

comparison with

baseline

condition where the surrounding

stimulus was statisnary. The results suggest that the ]aterstage where motion integration and

center-surround interactionappears, iscritical fordetermining motion deteetion _perfermance.

Key

words: visual motion, motion

integratjon,

illusion

Visual

rnotion

information

_passes through several

distinct

processing stages, inc]uding at

least

an

ear-lierlocalstage and a later_global stage where the

multiple

local

motion components are

integrated

(AdeLson

and Movshon, 1982;Rodman and Albright,

1989).

We

tested

how

vertical

illusory

motion affects

the detection sensitivity of horizontal _physical

mo-tionand

investigated

which stage of _processing

rep-resents the _perceptual limitof motion detcction.If

such a

limit

occurred at the earlier stage, the

detec-tion sensitlvity for horizontal rnotion wou]d not be

changed by

integration

with vertical compement.

Alternatively,

if

performance was

limited

at a

later

stage, thedetection sensitivity of the horizontal

mo-tioncould

be

affected

by

integration

with a vertical

component, To testthe_{possibilities,}we used an

i]]u-sion in which a central stationary stimulus appears

tomove inthe opposite direction

to

the surrounding

motion

(induced

motion;

Duncker,

1950).

And

we

examined the effect ef vertical induced motion on

perceived motion

direction

and

detection

sensitivity

for

central harizonta]motion, detection _{perforrnance.}

Experiment

1

Methods

The participants were 5 adults

(3

of whom were

* _{Department of Life Sciences,The University of}

Tokyo, 3L8-1 Komaba, Tokyo 153-8904

naive) with norma] or corrected-to-normahrision. As

shown

in

Figure

1,

the stimu]i consisted of a

Gabor

patch

(envelope

s.d. 2,58 deg) moving inleftward or

rightward surrounded

by

an annulus moving

in

up-ward er

downward

(inner

and outer

diameters

7.5

deg and 15 deg,respectively). For both stimuli, the

spatial frequency was O.53 cycles!deg.

The

fixation

point was

_provided

at 8 deg above the center of the

stimulus.

A

central

Gabo'r

patch was presented

for

500

ms. The participants were then asked toindicate

the_perceived motion

direction

o"he

Gabor

_patch

by

rotating an arrow-shaped visuaL icon using a mouse.

Results

When the surrounding stimulus was stationary',

the patch was perceived as moving horlzontally.

However, when thc sumrounding stimulus moved, the

patch

appeared tomove obliquely, biased toward the

direction

of the

induced

motion,

This

effect

became

more robust when the inducer moved faster

(4.42

deg/sec): when thisoccurred the central

_patch

ap-peared

to rnove alrnost vertically. Clearly,

the

in-duced motion and _physical motion were

integrated.

and

the

integrated

direction

d6pended

on the speed

of the inducer.

Experiment

2

Methods

The stimulus and experimental _protocol were

the

same as

for

Experiment

1.The participants were

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

184

The

Japanese

Journal

of

Psychonomic

Science

Vol.28,

No.

1

Figure !.

The

sttmu]us.

required tochoose the motion directionof the

Gabor

patch

'[rem

two alternatives,

left

or right.

The

tempo-ral frequency of theGabor patch was either O.047or

O.e24

deg/seq

and that of the

inducer

varied

from

O,

O.02,O.04,O.07,O,14,O.28,O.56,1.11,2.22to4.42degl

sec,

The

surrounding motion and

il]usory

vertical

motion were orthogonal tothe task and di,dnot help

judgment

of the

horizontal

direction.

Results

Figure

2

displays

the results of

'the

Experiment

2.

At

very

fast

surrounding motion speeds

(2.22,

4.42

degfsec), the motion detection sensitiv/ity was

de-graded, in cornparison to the control condition in

which the surrounding stimulus was stationary,

On

the other

hand,

at moderate surrounding motion

speeds

(from

O.14 to O.56deg/sec in fastercentral speeds: from O.07 to O.28deg/sec in s]ower central

speeds), the motion

detection

sensitivity was

en-hanced

in

comparison tothecontrol condition.

Experiment

3

To

elucidate whether

the

surrounding speed orthe

surrounding temporal

frequency

was important, we

carried out Experirnent

3

as a control experiment.

Methods

We

presented a counter-phase

fli'cker

stimulus

which contained thesame temporal frequency as the

surrounding stimu]us used

in

Experiment

2.

The

other _procedures were thesame as forExperiment 2.

Results

When

the surrounding temporal

frequency

was

higher,

thc

motion detectien sensitivity was

de-graded, which replicated theresults of Experiment

2.

However, an enhancement of motion detection

sensi-tivityat a lower surrounding temporal frequency was not observed.

as oxiff e,sgif e.7sg Q.1 ess as

um

wwww

tw$CewwsuuaIO.CpagTttagtigecip･--e-e

t

ttt

Cuatralwteept4rtegtsec**+' le..in:ml:---*ww

{te--- l"--l*t

-ntwhww-,itx.

dw'tsi"

scmecr=

ag

imwF

l

O Ctma {ne4 eaW7os4 Oms eSe 1,l{ 2.224.41 $wwund Velwhw {twsec)

"wnS,.,'ip""i.rkSwe,{fis."-Svcu}t)

Figure2. The results of Experirnent 2, The

curves indicate the correct response rates

under moving-surround conditions, and the

shaded lines

indicate

the correct rates under

stationary-surround conditions. Error bars

indicate =i 1

SEM.

Diseussion

The

degradation

effect at a

faster

sttrround can

be

exp]ained by the effect of the surrounding temporal

frequency,

as observed

in

a _previous study

(Takeuchi

& De Valois,200e). On the other hand, our results

revealed an enhancement effect at a slower surround

can

be

explained by the effect of induced motion, not

thetemporal

frequency

of the $urrounding stirnulus,

These

results suggest _that

later

stage processing,

in

which center-surround interaction and motion

inte-grationoccurs, iscritical for

determining

the percep-tua]limitof motion detection.

References

Adelson, E.H. & Movshon

J.

A.

_{1982}.

Phenomenal

coherence of moving visual _patterns.

Aigture,

3eO,

523-525.

Duncker, L.

(1950).

Uber induzierteBewegung, Ein

Beitrag zur

Theorie

optisch wahrgenommene

Bewegung,

In

W.

D. Ellis

{Ed.

& Trans.).

Source

Boola

of

Gestatt Rsychotogy. Kegan PauL Trench,

Trubner

&

Co.

pp.161-172.

{Original

work

lishedin

(1929).)

Rodman, H. R. & Albright,T.D,

(1989).

Single-unit

analysis of pattern-motion selective properties in

themiddle temporal vi,sual area

{MT).

EbePerimental

Brain

.Reseanrh,

75,53-64,

TakeuchL T. & De Valois K.K.

(2000)

Modulation of

perceived contrast by a moving surround, Vision

Research, 40,2697-2709.