Exogenously triggered perceptual switches in multistable ... Klanke - 2016 - Exogenously...Citation: Pastukhov, A., & Klanke,J. N. (2016). Exogenously triggered perceptual switches

Exogenously triggered perceptual switches in multistablestructure-from-motion occur in the absence of visualawareness

Alexander Pastukhov $

Department of General Psychology and MethodologyOtto-Friedrich-Universitat Bamberg

Bamberg GermanyCenter for Behavioral Brain Sciences

Magdeburg GermanyCognitive Biology Otto-von-Guericke Universitat

Magdeburg Germany

Jan-Nikolas Klanke $

Berlin School of Mind and BrainHumboldt-Universitat zu Berlin Berlin Germany

Center for Behavioral Brain SciencesMagdeburg Germany

Cognitive Biology Otto-von-Guericke UniversitatMagdeburg Germany

Here we characterize the duration of exogenouslytriggered perceptual switches in an ambiguously rotatingstructure-from-motion display and demonstrate theirindependence on visual awareness To this end wetriggered a perceptual reversal by inverting the on-screenmotion and systematically varied the posttriggerpresentation duration while collecting observersrsquo reportsabout the initial and final directions of illusory rotationWe demonstrate that for the structure-from-motiondisplay perceptual transitions are extremely brief (20ms) and can be considered instantaneous from anexperimental perspective We also report that althoughvery brief posttrigger intervals (10ndash20 ms) reliably initiatea perceptual reversal observers become aware ofperceptual switches only if the posttrigger presentationcontinues for at least 80 ms Additional experimentsdemonstrated that an observed lack of visual awarenessfor brief posttrigger presentation intervals cannot beattributed to either a systematic delay of visual awarenessor to backward masking Our results show thatexogenously triggered perceptual reversal can occur in theabsence of visual awareness extending earlier work onspontaneous reversals that indicated that neitherawareness nor attention may be required for multistableperception Methodologically the brevity and the shortlatency of induced perceptual reversals make themparticularly suitable for finely timed experiments such asmagnetoelectroencephalography studies

Introduction

Typically we experience our perception as stable andunambiguous in a sense that the same retinal inputresults in the same perception that remains constanteven during prolonged viewing However this seemingone-to-one relationship between sensory inputs andperception is an illusion (Gregory 2009 Metzger2009) This is particularly clear when it is violated byso-called multistable displays that are compatible withseveral distinct and comparably plausible perceptualinterpretations These displays force the visual percep-tion to continuously switch between alternatives despiteconstant sensory evidence (Blake amp Logothetis 2002Leopold amp Logothetis 1999)

The single most studied aspect of multistableperception is perceptual switching and we have a fairalthough hardly complete understanding of how theoccurrence of perceptual reversals can be predictedfrom the stimulus properties (Brouwer amp van Ee 2006Kang 2009 Levelt 1965) and prior perceptualexperience (Blake Westendorf amp Fox 1990 Kang ampBlake 2010 Nawrot amp Blake 1989 Pastukhov ampBraun 2011 Wolfe 1984) The neural correlates ofendogenous triggers of spontaneous reversals arecurrently debated but recent evidence from imaging

Citation Pastukhov A amp Klanke J N (2016) Exogenously triggered perceptual switches in multistable structure-from-motionoccur in the absence of visual awareness Journal of Vision 16(3)14 1ndash16 doi10116716314

Journal of Vision (2016) 16(3)14 1ndash16 1

doi 10 1167 16 3 14 ISSN 1534-7362Received September 14 2015 published February 12 2016

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 40 International LicenseDownloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

studies suggests that they are localized in sensory areasof the brain rather than in regions associated withexecutive control and attention (Frassle SommerJansen Naber amp Einhauser 2014 Knapen BrascampPearson van Ee amp Blake 2011 WeilnhammerLudwig Hesselmann amp Sterzer 2013) Less is knownabout the duration of perceptual reversals and aboutthe exact temporal relationship between a trigger eventchanges within a sensory representation and thefollowing visual awareness of that switch This isprimarily because we infer the timing of perceptualreversals from observersrsquo immediate responses whichare too variable to provide a reliable estimate(Pastukhov Vonau amp Braun 2012)

To overcome this limitation we investigated thetemporal characteristics of exogenously triggeredswitches (Pastukhov et al 2012 Treue AndersenAndo amp Hildreth 1995 see Figure 1 Movie 1) Toquantify the duration of exogenously triggered per-ceptual switches in structure-from-motion (SFM)displays we established the duration of the intermedi-atemixed perception following an exogenous triggerevent For this we report that perceptual reversals inSFM are extremely brief In addition we combinedseveral experimental measures to dissociate a domi-nance change within sensory representations and thevisual awareness of this change We demonstrate thatalthough the inversion of the on-screen motion appearsto trigger the reversal of perceptual dominance even ifthe posttrigger presentation is stopped after 20 ms theobservers become aware of that only if the followingpresentation period is at least 80 ms long In other

words exogenously triggered perceptual reversalsoccur in the absence of visual awareness

Methods

Observers

Procedures were in accordance with the Declarationof Helsinki and were approved by the medical ethicsboard of the Otto-von-Guericke Universitat Magde-burg lsquolsquoEthik-Komission der Otto-von-Guericke-Uni-versitat an der Medizinischen Fakultatrsquorsquo All participantshad normal or corrected-to-normal vision Apart fromthe second author observers were naive to the purposeof experiments and were paid for their participation

Apparatus

Stimuli were generated with MATLAB using thePsychophysics Toolbox (Brainard 1997) and displayedon a CRT screen (Iiyama VisionMaster Pro 514iiyamacom resolution 1600 3 1200 pixels refreshrate 100 Hz) The viewing distance was 73 cm so thateach pixel subtended approximately 00198 Observersresponded using a keyboard Background luminancewas kept at 36 cdm2 The experimental room was litdimly (ambient luminance at 80 cdm2)

Display

The SFM stimulus consisted of 50 dots distributedover the surface of the sphere The sphere diameter was578 and the dot diameter was 00578 For the mainobject (presented during the main Ton interval) thedots were distributed in such a way as to ensure aspecific distance between all left- and right-moving dotsat the time of the on-screen motion inversion (Ttriggeroffset of Tpreonset of Tpost presentation intervals) tomaximize the probability of triggering a perceptualswitch (see Stonkute Braun amp Pastukhov 2012 fordetails) For the probe stimulus (presented during theprobe interval) the dots were distributed randomlyover the surface of the sphere Both main and probestimuli were generated anew on every trial

Experiment 1

Nine observers (five of them female four male)participated in the experiment Each of the fourexperimental conditions (see below) was measured in aseparate experimental session Each session consisted

Figure 1 An exogenously triggered reversal of illusory rotation

in structure-from-motion displays Either an inversion of the 2D

motion at time Ttrigger (A illustrated for two example dots) may

trigger a reversal of the illusory rotation (B) or the illusory

rotation may remain stable following a spatial readjustment of

individual flow elements (C) See also Movie 1

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 2

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

of eight blocks and each block contained 70 trialsNote that the trials from Experiments 1 and 2 wereequally intermixed during each block (ie 35 trialsbelonged to Experiment 1 and 35 to Experiment 2)

Each trial consisted of a random onset delay (05ndash1s) a pretrigger interval (Tprefrac14 [500 625 750 875 1000]ms) an optional posttrigger interval (Tpostfrac14 [10 20 4080 160 320] ms) and a response interval (Figure 2A)The direction of the two-dimensional (2D) motion wasinversed at the onset of the posttrigger interval and thepresentation continued for a predefined amount of time(Tpost) The purpose of the on-screen motion inversionwas to trigger a reversal of the perceived illusoryrotation (Pastukhov et al 2012 Treue et al 1995 seeFigure 1) The lsquolsquono inversionrsquorsquo presentation conditioncontained no on-screen motion inversion and corre-spondingly no postinversion presentation interval(Figure 2B)

The dots were distributed on the surface of theillusory sphere in such a way as to ensure a specificminimal distance between pairs of left- and right-moving dots at the time of the on-screen motioninversion (Stonkute et al 2012) We used four interpairdistances to systematically manipulate the strength ofthe motion transient and therefore the probability ofsuccessfully induced perceptual reversals The fourconditions were labeled according to the strength of themotion transient Strong (1)S1 Strong (2)S2MediumM and WeakW The maximal induceddestabilization was determined using the longest post-trigger interval duration (Tpost frac14 320 ms for furtherdetails see the Results sections for Experiments 1 and2) Please note that the effectiveness of the motiontransient in inducing a perceptual reversal depends notonly on its strength but also on a prior perceptualhistory (Pastukhov Vivian-Griffiths amp Braun 2015)The same procedure but using variable pretriggerintervals has been used to study the onset perception ofSFM displays (Pastukhov 2015)

Observers reported on initial (beginning of Tprelabeled as R(1)) and final (end of Tpost interval or endof Tpre interval for lsquolsquono inversionrsquorsquo condition labeled asR(2)) directions of illusory rotation Observers had anoption of reporting an unclearmixed percept via thelsquolsquodownrsquorsquo key The response times were measured withrespect to the end of the presentation (offset of Tpost orTpre for lsquolsquono inversionrsquorsquo condition) Accordinglyperceptual destabilization due to an endogenous triggerevent was quantified as

Preversal frac14 P

Reth1THORN 6frac14 Reth2THORN eth1THORN

Group averages were fitted with a logistic functionusing the Palamedes toolbox (Prins amp Kingdom 2009)Standard errors of measurement for the free parameters

of the logistic function were obtained using a bootstrapprocedure implemented in the Palamedes toolbox

Experiment 2

Nine observers (five female four male) participatedin the experiment The procedure was identical to thatof Experiment 1 but for an additional probe SFMdisplay The visual sequence of Experiment 1 wasfollowed by a brief blank interval (Tblankfrac14 50 ms) andthe probe SFM display (Tprobe frac14 500 ms) The probestimulus was a different sphere (ie the location ofindividual flow elements was different from that of themain sphere) Observers reported on the initial rotationof the main stimulus and on the final direction ofillusory rotation of the probe display See Figure 3ANote that trials from Experiments 1 and 2 were equallyintermixed during each block (ie 35 trials belonged toExperiment 1 and 35 to Experiment 2)

Experiment 3

Six observers (three of female three male) partici-pated in the experiment The SFM display was identicalto the Strong (1)S1 condition of Experiments 1 and 2The presentation schedule of the SFM display wassimilar to that of Experiments 1 and 2 but without thelsquolsquono inversionrsquorsquo condition and with only long post-trigger intervals (Tpost frac14 [140 150 160 170 180] ms)The SFM display was accompanied by a yellow dot(diameter 0758) that moved clockwise along thecircular trajectory (radius 578) with a speed of 6008sThe initial position of the dot was randomized Pleasesee Movie 2 which demonstrates several presentationtrials without a response interval (please note that theactual experimental display looked different because ofa higher refresh rate) After the presentation the SFMdisplay was taken of the screen and the dot was movedto a random location that was at least 458 away fromthe location the dot was in at the time of the on-screenmotion reversal The observers were instructed tomemorize the location of the yellow dot at the time ofthe illusory rotation reversal They used arrow keys(left and right) to move the dot to the memorizedlocation and lsquolsquoEnterrsquorsquo to confirm it Observers had anoption to report the lack of reversal using a lsquolsquoQrsquorsquo key(116 6 127 of trials)

Experiment 4

Seven observers (four female three male) partici-pated in the experiment Apart from an added 500-msdelay before the response prompt the procedure was

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 3

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

Figure 2 Experiment 1 (A) Schematic procedure Each trial consisted of a random-onset delay (Tdelayfrac14 05ndash1 s) a presentation

interval (Tonfrac14 Tprethorn Tpost) and a response interval The direction of the on-screen motion was inversed at the end of the Tpre interval

and the presentation continued for a predefined amount of time (Tpost) The purpose of the on-screen motion inversion was to trigger

a reversal of the perceived illusory rotation Observers reported on the initial (R1) and final (R2) directions of illusory rotation (B)

Schematic procedure for the lsquolsquono inversionrsquorsquo presentation condition The procedure was similar but for the omitted on-screen motion

inversion and the lack of the posttrigger interval (C) Probability of reversal as a function of the posttrigger interval duration Tpost(mean and 95 confidence interval based on binomial distribution) Curves depict best-fitting logistic functions (D) Distributions of

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 4

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

Figure 3 Experiment 2 (A) Schematic procedure The procedure was similar to that of Experiment 1 but for a blank interval and a

probe display which followed the presentation of the main stimulus (Tblankfrac14 50 ms Tprobefrac14 500 ms both marked by orange color)

The probe display was a different ambiguously rotating SFM sphere Observers reported first on the initial direction of rotation of the

main stimulus R1 and then on the final direction of rotation of the probe stimulus R2 (BndashE) Probability of a perceptual switch as the

function of the postinversion interval Tpost Filled circles results for Experiment 2 Open circles results for Experiment 1 replotted for

comparison Asterisks mark statistically significant differences between the two experiments (paired-sample t test a Bonferroni

correction for multiple testing) Subfigures show data for (B) Strong (1) (C) Strong (2) (D) Medium and (E) Weak conditions

threshold (a) and slope (b) parameters obtained by parametric bootstrapping (1000 iterations) Solid curves encircle 6827 of the

data points Bars depict the mean and standard error for each distribution and parameter (E) Normalized response time for the initial

direction of rotation (RT1) as a function of the posttrigger interval duration (Tpost) Gray stripe SEM for an overall average (F) Fraction

of lsquolsquounclearrsquorsquo responses for the final direction of rotation (R2) as a function of the posttrigger interval duration (Tpost) Gray stripe SEM

for an overall average

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 5

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

identical to that of Experiments 1 and 2 Both types oftrials (with and without probe stimulus) were randomlymixed within a block

Experiment 5

Nine observers (five female four male) participatedin the experiment The procedure was similar toExperiment 2 Two conditions were used lsquolsquono inver-sionrsquorsquo and Tpost frac14 20 ms labeled here as lsquolsquowithinversionrsquorsquo The blank duration was systematicallyvaried Tblank frac14 [50 100 200 400 800] ms

Results

Experiment 1 Time necessary for the visualawareness of the perceptual reversal

In the first experiment we sought to estimate thetime interval between the trigger event and the displayoffset that is necessary for a perceptual reversal andorfor the visual awareness of it To this end we reversedthe on-screen motion of all flow elements at apredefined moment of time (Ttrigger in Figure 2A) whilesystematically varying the duration of a posttriggerinterval Observers reported on the initial and finaldirections of illusory rotation for each presentationinterval

To confirm that the observers faithfully and consis-tently reported their subjective perception of illusoryrotation our experimental design contained twocontrol conditions lsquolsquono inversionrsquorsquo and Tpostfrac14 320 msThe lsquolsquono inversionrsquorsquo condition is simply a brief andunperturbed presentation of an SFM display (Figure2B) which should lead to stable illusory rotation withinthe presentation interval Conforming our expecta-tions the observers tended to report the same directionof rotation at the beginning and at the end of thepresentation (No inversion in Figure 2C) Preversal (S1lsquolsquono inversionrsquorsquo) frac14 006 [003ndash008] (mean and 95confidence interval for binomial distribution) Preversal

(S2 lsquolsquono inversionrsquorsquo)frac14 003 [001ndash005] Preversal (M lsquolsquono

inversionrsquorsquo) frac14 001 [0005ndash03] and Preversal (W lsquolsquonoinversionrsquorsquo) frac14 002 [001ndash004]

Conversely Tpost frac14 320 ms was the longest presen-tation interval which provided observers with the bestopportunity to observe and report a reversal of illusoryrotation (Tpost frac14 320 ms in Figure 2C) In agreementwith prior work (Stonkute et al 2012) a strongermotion transient due to the on-screen motion inversionproduced more frequent switches of illusory rotationPreversal (S1 320 ms)frac14 084 [08ndash088] Preversal (S2 320ms)frac14 085 [081ndash089) Preversal (M 320 ms)frac14 074 [07ndash079] and Preversal (W 320 ms) frac14 042 [037ndash048]

The time intervals in between these two extremesrepresent a growing probability of reported perceptualswitches (Figure 2C) Group averages across the nineobservers were fitted with a logistic function For allfour conditions the posttrigger duration that led tothreshold reports of visual awareness of perceptualreversals was approximately 65 to 75 ms Perceptualswitches were reliably (Preversal 099Preversal [320 ms])reported 120 to 150 ms after the trigger event (see Table1 Figure 2D)

Because perceptual adaptation has a strong influenceon the perception of multi-stable displays (Blake et al1990 Kang amp Blake 2010 Pastukhov amp Braun 2011van Ee 2009) we analyzed its effect on inducedperceptual reversals in the current study First weexamined the effect of the short-term adaptation bycomparing the probability of reversals for the shortest(Tpre frac14 500 ms) and the longest (Tpre frac14 1000 ms)pretrigger intervals but found no significant changet(251) frac1416 p frac14 01 paired-sample t test Next werepeated the same analysis but for trials from the firsthalf versus trials from the second half of eachexperimental session to assess the influence of the long-term adaptation Here we found a small but significanteffect of adaptation t(251)frac143 p frac14 0003 Howeverlong-term adaptation influenced only an overall prob-ability of induced perceptual reversals (ie guess andlapse rate) but not the threshold or the slope of thepsychometric functions (data not shown the effect wasweaker than but qualitatively similar to the oneillustrated in Figure 4) Accordingly we found noevidence that for the paradigm used here either short-term or long-term adaptation alters the speed ofinduced perceptual reversals (see also Experiment 2)

Strong0 Strong Medium Weak

Threshold [ms] 642 6 105 726 6 106 751 6 107 760 6 110

Slope 261 6 029 241 6 026 245 6 030 273 6 053

Support [ms] 115 1563 1283 907

Guess rate 0026 6 001 0028 6 001 0007 6 192 0014 6 010

Lapse rate 014 6 002 012 6 003 025 6 024 056 6 003

Table 1 Experiment 1 summary of logistic function fits

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 6

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

As an additional measure we have analyzed the effectof both condition and posttrigger interval duration onresponse time and the fraction of mixed reports (seeTable 2) The response times for both intervals (markedas RT1 and RT2 in Figure 2A) were affected by thecondition and the posttrigger interval duration How-ever their effect was bigger on the first response interval

and therefore we concentrated on RT1 in the analysisbelow With respect to mixed perception responses wefound a marginally significant effect of the posttriggerinterval and a significant interaction of the effects for thesecond response interval when the observers respondedabout the final direction of illusory rotation Theweakness of both effects is likely to be explained by a

Figure 4 Experiment 2 effect of the Tpre interval duration (A) Probability of the perceptual switch as a function of the posttrigger

Tpost interval duration (BndashE) Comparison between the shortest (Tprefrac14500 ms downward-pointing triangles) and longest (Tprefrac141000

ms upward- pointing triangles) preinversion intervals Subfigures show data for (B) Strong (1) (C) Strong (2) (D) Medium and (E)

Weak conditions

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 7

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

very low overall fraction of mixed reports (18 6 09of all final direction reports across all conditions and allpostinterval durations) More specifically the thresholdduration was associated with longer response times(Figure 2E RT1[lsquolsquono inversionrsquorsquo]frac14 5626 6 87 msRT1[80 ms]frac14 610 6 106 ms) t(35)frac1435 pfrac14 0001paired-samples t test and a higher fraction of mixedpercepts (Figure 2F mixed[lsquolsquono inversionrsquorsquo]frac14 002[0017ndash0025] [mean and 95 confidence interval forbinomial distribution] mixed[80 ms]frac14 006 [0046ndash0072]) t(35)frac1424 pfrac14 0023 paired-samples t test Inother words the observers were slower to respond andwere less certain about the final direction of illusoryrotation for threshold duration displays However a lownumber of mixed reports indicates that mixed phaseswere very brief and were rarely perceived even under themost favorable threshold conditions (Tpostfrac14 80 ms)

Next we quantified how the motion transientrsquosstrength altered the speed of induced perceptualreversals (ie threshold andor slope of a psychometricfunction) Comparing the two most different conditions(Strong [1] vs Weak) we found no significantdifferences for the guess rate (p frac14 098 statisticalcomparison using Monte Carlo method by Palamedestoolbox) or for the slope (p frac14 055) parametersHowever there was a highly significant difference in thelapse rate (p 0001) and a significant difference in thethreshold parameters (pfrac14 00183 see also Figure 2D)Thus not only did a larger pairing distance producemore frequent perceptual switches but those switcheswere also perceived slightly earlier (see Table 1)

A critical aspect of the data which would serve for animportant comparison in Experiment 2 is the perceptionof illusory rotation for very brief posttrigger intervals(Tpostfrac14 [10 20] ms) Not only have the observersconsistently reported the same direction of rotation atthe beginning and at the end of the presentation (Figure2C) but they were also very fast to respond (Figure 2E)even faster than on trials without an on-screen motion

inversion RT1(lsquolsquono inversionrsquorsquo)frac14 5626 6 87 ms (mean6 standard deviation) RT1(20 ms)frac14 5362 6 67 mst(35)frac14 27 pfrac14 001 (paired-samples t test) In additionthey reported very few mixed percepts (Figure 2F)Thus all three measures (subjective reports of clearperception response times and subjective reports onmixed perception) indicate that there was no perceptualdifference between trials with a very brief postintervalduration (Tpostfrac14 [10 20] ms) and trials without stimulusperturbation (lsquolsquono inversionrsquorsquo)

Experiment 2 Probing an interruptedperceptual switch

Results of Experiment 1 demonstrated that a reversalof illusory rotation was perceived only if the displaypresentation continued for another 80 to 150 ms afterthe exogenous trigger (the on-screen motion inversion)The perceptual switch itself was a very brief event asmanifested by a very low fraction of mixed perceptseven for threshold conditions (see Figure 2F) Thenecessity for this prolonged stimulation may come fromtwo sources First this time may be required for acomputation of an altered on-screen motion which inturn triggers a very brief perceptual switch This wouldmean that for very brief postinversion intervals (Tpostfrac14[10 20] ms) the reversal of an illusory rotation was notperceived because the sensory representation ofillusory rotation remained stable at that time point anda longer posttrigger presentation is required to trigger areversal within them Second the subjective awarenessof a new direction of illusory rotation may have beenimpeded by an earlier perception for example viaforward or backward masking (Enns amp Di Lollo 2000)In this case a reversal of illusory rotation withinsensory representations may have occurred soon afterthe on-screen motion inversion (eg already after 10-to 40-ms presentation) and the prolonged presentationwould be required only to overcome masking

To distinguish between these two possibilities wemodified Experiment 1 by appending the displaysequence of Experiment 1 with a brief blank (Tblankfrac14 50ms) and a probe stimulus (Figure 3A additional blankand probe intervals are marked with orange) The probewas presented for 500 ms giving enough time for theobservers to become aware of its direction of rotationrendering forward masking irrelevant The probestimulus was a different sphere (ie the location ofindividual flow elements was different from that of themain sphere) This change interrupted the continuity ofthe on-screen motion as well as the continuity of 3Drepresentations of individual dots This way only therepresentation of an interpolated 3D object couldremain stable Accordingly we assumed that because ofa very brief interruption the illusory rotation of the

Factor

Condition

(df 324)

Tpost(df 648)

Condition

3 Tpost(df 18144)

F p F p F p

RT1 166 0001 118 0001 14 011

RT2 55 0005 38 0004 075 075

Unclear

perception R1 17 02 12 03 09 057

Unclear

perception R2 2 014 22 0059 19 0019

Table 2 Experiment 1 results for the repeated-measuresanalysis of variance for response time and mixed reports NotesBold font marks statistically significant effects

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 8

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

interpolated object should persist (Pastukhov amp Braun2013) and the direction of rotation of the probe will berepresentative of the final direction of illusory rotationof the main (original) sphere (see Experiment 5 for aconfirmation of this assumption) In all other respectsthe procedure was identical to that of Experiment 1 Tofacilitate the comparison trials from Experiments 1 and2 were mixed together within a single block of anexperimental session (see Methods for details)

To summarize our experimental procedure limitedthe time for 2D motion extraction interrupted persis-tence at the level of individual flow elements preservedpersistence at the level of an interpolated object andgave enough time for the visual awareness of a newdirection of illusory rotation to emerge

Two postinversion interval durations were used as acontrol lsquolsquono inversionrsquorsquo and Tpost frac14 320 ms As inExperiment 1 the longest postinversion interval wasexpected to reveal the highest fraction of successfullytriggered switches and we found no difference betweenthe two experiments (see Tpost frac14 320 ms in Figure 3Bthrough E all p values 005) This means that thefinal direction of illusory rotation of the probe stimuluswas a reliable indicator of the final direction of illusoryrotation of the main sphere (see also Experiment 5)

For the lsquolsquono inversionrsquorsquo presentation condition therewas no on-screen motion inversion so the perceptualswitches were not exogenously triggered during thepresentation of the main sphere Therefore as inExperiment 1 we were expecting the same perceptionof illusory rotation to be reported for both displaysHowever we found that for all four conditions theprobability of the switch was significantly higher inExperiment 2 than in Experiment 1 (all p values below001 paired-samples t test see lsquolsquono inversionrsquorsquo in Figure3B through E results of Experiment 2 are marked byfilled circles and results of Experiment 1 are marked byopen circles and are replotted as a comparison) Preversal

(S1 0 ms)frac14 027 [022ndash032] (mean and 95 confidenceinterval for binomial distribution) Preversal (S2 0 ms)frac14036 [03ndash041] Preversal (M 0 ms)frac14 024 [02ndash03] andPreversal (W 0 ms) frac14 018 [014ndash023] This mildperceptual destabilization is typical for briefly inter-rupted multistable displays (Kornmeier Ehm Bigalkeamp Bach 2007 Orbach Ehrlich amp Heath 1963Pastukhov amp Braun 2013) and is likely to reflect anaccumulated perceptual adaptationfatigue To confirmthis and analogously to the analysis we performed forExperiment 1 we assessed the effects of the short-termand long-term adaptation (respectively the effect of thepreinversion interval duration Tpre and the differencebetween trials from the first half of an experimentalsession vs trials from the second half) The effect ofboth short-term adaptation t(251)frac14101 p 0001paired-sample t test for Tprefrac14500 ms versus Tprefrac141000ms (see Figure 4) and long-term adaptation t(251) frac14

46 p 0001 were highly significant However inboth cases adaptation shifted the entire psychometriccurve vertically but not horizontally (see Figure 4) Aswith Experiment 1 this indicates that althoughaccumulated adaptation significantly increases theprobability of endogenously triggered perceptual re-versals it has little or no influence on their duration

In contrast to Experiment 1 we were unable to fitgroup averages with a logistic function The reason forthis was that the probability of reversal reached itsmaximum already after Tpostfrac14 20 ms for all fourexperimental conditions The probability of reversal wassignificantly different between the two experiments for allpostinversion intervals shorter than 160 ms (paired-samples t test with a Bonferroni correction for multipletests statistical significance is indicated by stars in Figure3B though E) Note that in Experiment 1 the sameposttrigger intervals (Tpostfrac14 [10 20] ms) were perceptu-ally indistinguishable from the lsquolsquono inversionrsquorsquo condition

With respect to the research question we formulatedfor this experiment this means that the on-screenmotion extraction is complete and the reversal ofperceptual dominance within sensory representations isinitiated after 10 to 20 ms of the posttrigger presenta-tion This is particularly evident in the Weak condition(Figure 3E) where the fraction of reported switchesreaches the maximal level for the same Weak conditionof Experiment 1 already for the 20-ms posttriggerinterval and remains at this level for all longerposttrigger intervals In other words our probeparadigm reveals exactly as many switches for briefposttrigger intervals of Experiment 2 as for the longestposttrigger intervals in Experiment 1 Accordingly inExperiment 1 the lack of visual awareness of thereversal occurs despite a dominance change within thesensory representation of illusory rotation

Because for all four conditions the maximal destabi-lization was reached already after the 20-ms posttriggerpresentation we were unable to assess the influence ofthe motion transientrsquos strength on the speed (orconversely duration) of induced perceptual reversalsThis indicates that a small but significant change in thethreshold between the Strong (1) and Weak conditionsobserved in Experiment 1 most likely was not due to afaster perceptual reversal within a sensory representa-tion Instead a stronger motion transient might havefacilitated a faster propagation of this reversal intovisual awareness perhaps by better attracting attention

Experiment 3 Estimated time of visualawareness of the illusory rotation reversalduring prolonged presentation

Although a change in the perceptual dominancewithin the sensory representations is initiated shortly

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 9

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

after the trigger event (see Experiment 2) it is possiblethat the visual awareness of that change consistentlylags in time (Libet 1999) In other words the visualawareness of an exogenously triggered reversal consis-tently occurs not earlier than at least 40 to 60 ms afterthe trigger event Therefore we asked observer toestimate the time when they perceived the reversal andexamined whether these estimates significantly andconsistently lagged behind the exogenous trigger event

To this end we adopted a lsquolsquoLibetrsquos dotrsquorsquo paradigm(Libet Gleason Wright amp Pearl 1983) that waspreviously used to estimate the time of spontaneousreversals of illusory rotation (Pastukhov et al 2012)An ambiguously rotating sphere was accompanied by ayellow dot that was circling around the SFM displayThe observers were instructed to memorize the locationof the dot at the time when they perceived a reversal inthe illusory rotation (see Movie 2) During the laterresponse interval they moved the dot to the memorizedlocation thus allowing us to estimate the time of theperceived reversal The initial location of the yellow dotwas randomized and its location at any given time wasnot informative about the time of the trigger event

Empirical cumulative densities functions of theestimated time of induced perceptual reversals for sixobservers are plotted in Figure 5 (black curves and lefty-axis mean estimated event times relative to thetrigger event are marked by teal color) Although theobservers varied in the mean estimated time of theinduced perceptual reversal they showed no systematicbias The group average estimated time of the reversalwas 15 6 21 ms and it was not significantly differentfrom zero t(5) frac14 007 p frac14 094 In other words wefound no tendency to perceive induced perceptualreversals as occurring significantly later than thephysical trigger event Importantly the same sixobservers required at least 80 ms of continued visualpresentation to develop a visual awareness of aninduced switch in Experiment 1 (see red lines and righty-axis in Figure 5 replotted for comparison purposes)Therefore we conclude that the results of Experiments1 and 2 cannot be explained by a systematic delaybetween the time of the exogenous trigger event and thesubjective perception of an illusory rotation reversal

Experiment 4 Lack of awareness is not due tobackward masking

Our results of Experiments 1 and 2 show thatalthough the switch in perceptual dominance appearsto be initiated if the presentation continues for 20 msafter the on-screen motion inversion (Experiment 2)the brief presentation times preclude the visualawareness of that switch (Experiment 1) One possibleexplanation for this dissociation between a sensory

change and its visual awareness is masking (Enns amp DiLollo 2000) For example the earlier perception of anopposite direction of motion could produce forwardmasking Alternatively it is possible that the responseprompt which appeared on the screen immediatelyafter the main display produced backward masking

Here we tested the latter hypothesis by replicatingthe lsquolsquoStrong (1)rsquorsquo condition from Experiment 1 and 2(marked respectively with open and filled circles inFigure 6) but with an additional 500-ms blank intervalinserted before the response prompt To replicateExperiment 1 it was presented after the main SFMdisplay (see also Figure 2A) and the stimulus onsetasynchrony was increased from 20 to 320 ms to 520 to820 ms For the replication of Experiment 2 a blankwas inserted after the presentation of the probe display(see also Figure 3A)

The results of Experiment 4 are presented in Figure6 If a delay of visual awareness of the perceptualreversal in Experiment 1 was due to backward maskingthe additional blank should have attenuated its effectTherefore the number of reported perceptual reversalsfor brief Tpost durations for the replication ofExperiment 1 should have increased and the filled-circles curve in Figure 6 should have become similar tothe open-circles curve (curves correspond respectivelyto replications of Experiments 1 and 2) However thecurves in Figure 6 both qualitatively and quantitativelymatch the results of Experiments 1 and 2 Specificallyin replication of Experiment 1 short postinversionintervals (Tpost 40 ms) lead to consistent reports ofperceptual stability We conclude that the lack ofawareness is not explained by backward masking fromthe response prompt

Experiment 5 The direction of illusory rotationin the probe display reflects the most recentperceptual state before the interruption

The experimental procedure for the Experiment 2was based on the assumption that the direction ofillusory rotation of the probe displays reflects the mostrecent (final) state of the main display before theinterruption This assumption was based on earlierwork that showed that for brief blank intervals theperceptual dominance is stabilized by neural persis-tencemdasha quickly decaying activity of an originallydominant neural population (Pastukhov amp Braun2013) However it is possible that the perceptualdominance of the probe display reflected anotherhistory effect such as a sensory memory of multistabledisplays (Adams 1954 Leopold Wilke Maier ampLogothetis 2002 Orbach et al 1963 Ramachandranamp Anstis 1983)

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 10

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

To control for this possibility we replicated Exper-iment 2 using only two conditions lsquolsquono inversionrsquorsquo andTpost frac14 20 ms (labeled here as lsquolsquowith inversionrsquorsquo) butwith a broad range of the blank interval durations

(Tblank frac14 [50 100 200 400 800] ms) The purpose ofthe latter was to dissociate the influence of two historyeffects in question Whereas neural persistence decayswithin 400 to 500 ms (Pastukhov amp Braun 2013)

Figure 5 Estimated time of an illusory rotation reversal individual observers Black color and left vertical axis Empirical cumulative

densities function of an estimated time when a perceptual reversal occurred relative to the time of the trigger event (Ttrigger) CDFfrac1405 corresponds to the mean estimated time of a perceptual switch (marked by teal color) Although the accuracy and the bias of an

estimated switch time vary between individual observers we found no systematic tendency to perceive an exogenously triggered

switch to occur significantly later than the trigger event Red color and right vertical axis Results of the Strong (1) condition of

Experiment 1 replotted for comparison In contrast to Experiment 3 all six observers were very similar in that they required 80 ms

of continued visual presentation to become aware of an induced perceptual reversal

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 11

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

sensory memory is characterized by very long decaytimes of dozens of seconds or even minutes (Leopold etal 2002) Accordingly if the perceptual dominance ofthe probe displays was determined by sensory memorythe blank interval duration should have a minimaleffect Conversely if illusory rotation was stabilized byneural persistence this effect should be absent forblank intervals longer than 400 to 500 ms

The results of Experiment 5 are plotted in Figure 7The lsquolsquono inversionrsquorsquo condition (gray filled circles inFigure 7) served as a baseline to determine the effect ofthe blank interval duration on the probability ofperceptual reversals in the absence of exogenouslytriggered perceptual switches The results revealed aninverted U-shape and were qualitatively consistent withprevious reports (Kornmeier et al 2007 OrbachEhrlich amp Vainstein 1963 Pastukhov amp Braun 2013)For the lsquolsquowith inversionrsquorsquo condition (orange open circlesin Figure 7) the Tblank frac14 50 ms duration was identicalto that of Experiment 2 and replicated a reliableswitching effect of the prior on-screen motion inver-sion Critically this effect disappeared for blankslonger than 100 ms This short-lived effect is consistentwith neural persistencehysteresis but not with sensorymemory of multistable displays or any other bias thatoperates at the time scale of seconds We conclude thatthe perceptual dominance of the probe displays reflectsthe latest perceptual state of a prior display

Discussion

Here we investigated the perception of an exoge-nously triggered reversal of illusory rotation in SFMdisplays We report that the reversals themselves arevery brief (Experiment 1) and that the change indominance in the sensory representation is initiatedshortly after the trigger event as even a 20-msposttrigger presentation duration is sufficient for this(Experiment 2) However the observers become awareof that switch only if the presentation continues for atleast 80 ms after the trigger event (Experiment 1) Thiseffect cannot be explained either by a systematic delayof visual awareness (Experiment 3) or by backwardmasking (Experiment 4) Therefore we conclude thatexogenously triggered reversals are brief and can occurin the absence of visual awareness

Induced perceptual switches occur in theabsence of visual awareness

The results of Experiments 1 and 2 demonstrate thatalthough a reversal of the perceptual dominance withinthe sensory representations of illusory rotation isinitiated within 20 ms after the trigger event theobservers become aware of that only if the presentationcontinues for at least 80 ms If the presentation iscurtailed using shorter posttrigger intervals (Tpostfrac14 10ndash20 ms) the observers fail to notice the reversal and

Figure 6 Experiment 4 An additional blank before the response

interval minimizes the masking but has no effect on perceptual

reversals Probability of a perceptual switch is plotted as a

function of the postinversion interval Tpost Filled circles

replication of Experiment 2 Open circles replication of

Experiment 1 Asterisks mark the statistically significant

differences between the two conditions (paired-sample t test

the Bonferroni correction for multiple testing)

Figure 7 Experiment 5 Probability of the perceptual switch as a

function of the blank interval for trials with (open gray circles)

and without (filled orange circles) an on-screen motion

inversion The influence of the on-screen motion inversion was

significant only for very short blank intervals (Tblank 200 ms)

Asterisks mark statistically significant differences between

lsquolsquowith inversionrsquorsquo and lsquolsquono inversionrsquorsquo conditions (paired-samples

t test the Bonferroni correction for multiple testing)

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 12

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

both their responses and presumably perception arequalitatively and quantitatively similar to that of anunperturbed stable illusory rotation (see Experiment 1)Earlier work showed that exogenously triggeredreversals also occur in the (near) absence of attention(Stonkute et al 2012) Therefore we can conclude thatneither awareness nor attention is necessary forexogenously triggered reversals of perceptual domi-nance in SFM

Our results raise further questions about thecontribution of top-down factors such as attention andvisual awareness to perceptual switches Althoughshifts of attention were postulated as a possiblemechanism behind perceptual reversals (Leopold ampLogothetis 1999) and their causal effect on multistableperception is well documented (Brouwer amp van Ee2006 Chong Tadin amp Blake 2005 Mitchell Stoner ampReynolds 2004) later work demonstrated that atten-tion may not be required for spontaneous reversals(Pastukhov amp Braun 2007 Roeber Veser Schroger ampOrsquoShea 2011 but see Brascamp amp Blake 2012 ZhangJamison Engel He amp He 2011) Similarly not onlyexogenously triggered but also endogenously triggered(spontaneous) reversals can occur without visualawareness (Brascamp Blake amp Knapen 2015 Plato-nov amp Goossens 2014)

The similarity between our results and the findingson endogenously triggered reversals indicates thatneural populations responsible for initiation of spon-taneous perceptual switches are likely to be located insensory regions of the brain This idea fits well withprior psychophysical experiments (Alais Cass OrsquoSheaamp Blake 2010 Brascamp van Ee Noest Jacobs amp vanden Berg 2006 van Ee 2009) modeling (Noest vanEe Nijs amp van Wezel 2007 Shpiro Moreno-BoteRubin amp Rinzel 2009) and recent imaging studies(Brascamp et al 2015 Frassle et al 2014 Knapen etal 2011) Accordingly it strengthens the idea thatalthough both awareness and attention modulatemultistable perception and the occurrence of perceptualreversals they are not causally responsible for them

Exogenously versus endogenously triggeredperceptual reversals

The presented study used exogenously triggeredperceptual reversals and this warrants a questionabout how much our findings can tell us about generalmechanisms behind multistable perception and spon-taneous perceptual switches However one mustremember that a spontaneous reversal can be triggeredby internal forces as divergent as an involuntary eyemovement an intrinsic neural noise or a shift ofattention Accordingly even if a conceptual differencebetween exogenously and endogenously triggered

perceptual reversals is clear it is less obvious howindividual endogenous triggers differ from exogenousones in terms of events occurring at the level of neuralrepresentations

For example although a multistable display itselfmay remain constant throughout the entire presenta-tion its retinal image never does Even when theobservers are faithfully fixating the retinal image of adisplay is constantly changing because of eye tremordrift and microsaccades (Martinez-Conde Macknik ampHubel 2004) These changes of the retinal image areendogenously generated but can trigger a perceptualreversal (van Dam amp van Ee 2006) just like exogenouschanges due to an inversion of the on-screen motion ordue to a brief change of an image contrast (KimGrabowecky amp Suzuki 2006)

Furthermore all neural representations startingalready at the retinal level are intrinsically noisy(Faisal Selen amp Wolpert 2008) This means that allneural representations involved in multistable percep-tion both ones that can be considered lsquolsquoinputsrsquorsquo inmodeling terms and those that correspond to compet-ing percepts undergo constant random changes Thesenoise-driven fluctuations in neural representations arecurrently thought to be the main source of spontaneousreversals based on both experimental (Brascamp et al2006 Pastukhov amp Braun 2011 van Ee 2009) andmodeling (Moreno-Bote Rinzel amp Rubin 2007 Noestet al 2007 Pastukhov et al 2013) perspectives Just asan inversion of the on-screen motion these noise-induced fluctuations are transient and do not produce along-lasting bias in favor of a particular perceptionAnd similar to the on-screen motion inversions theymay trigger a reversal in the perceptual dominancewhen they occur at an appropriate moment (Moreno-Bote et al 2007 Noest et al 2007) However just likeexogenous triggers they may also be ignored by thevisual system manifesting themselves as brief periodsof destabilization (Brascamp et al 2006 NaberFrassle amp Einhauser 2011 Pastukhov amp Braun 2011)

Taken together this suggests that the differencebetween the endogenous noise-driven transient changesin neural representations and the exogenous-driventransient changes in neural representations may be of aquantitative rather than qualitative nature Accord-ingly the interpretation of our results and of otherwork on the exogenously triggered reversals will befacilitated primarily by a better understanding ofdivergent endogenous causes of spontaneous reversals

Duration of perceptual switches

Experiment 1 demonstrated that in agreement withcurrent models of multistable perception (Laing ampChow 2002 Moreno-Bote et al 2007 Moreno-Bote

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 13

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

Knill amp Pouget 2011 Noest et al 2007) the durationof perceptual switches in SFM is extremely brief Theobservers rarely reported unclear perception evenunder the most favorable threshold conditions (80 msin Experiment 1) These nearly instantaneous switchesmay appear to be drastically faster than longer andeasily noticeable transitions between two clear perceptsin binocular rivalry (Blake OrsquoShea amp Mueller 1992Brascamp et al 2006 Pastukhov amp Braun 2011)However it is possible that the difference lies not in thenature of two multistable displays but in the exactdefinitions of perceptual switches and perceptualtransitions

For binocular rivalry a perceptual transition can bedefined as a perception that is different from the stateof exclusive visibility and typically includes piecemealrivalry as well as episodes of binocular fusionAlthough both of these perceptual states clearly differfrom exclusive visibility they also do not correspondto the perceptual switch Binocular fusion is a defaultand different state of binocular vision (Wolfe 1983)The piecemeal rivalry is the patchy appearance whensome spatial regions are dominated by one eye whereasother regions are dominated by the other eye It ismore likely to occur for bigger visual displays (Blake etal 1992 Kang 2009 OrsquoShea Sims amp Govan 1997)and is reduced in the presence of additional groupingfactors such as rotation (Haynes Deichmann amp Rees2005) Accordingly although one can talk aboutlsquolsquomixed perceptionrsquorsquo with respect to the entire imageindividual patches are in exclusive visibility statesAccordingly although periods of nonexclusive visibil-ity may be long they do not necessarily correspond totransient reversals of perceptual dominance postulatedin current models of multistable perception (Laing ampChow 2002 Moreno-Bote et al 2011 2007 Noest etal 2007)

To summarize perceptual reversals in SFM areextremely brief and it is for future research todetermine whether the same is true for perceptualswitches in binocular rivalry and other multistabledisplays However it will be important to distinguishbetween perceptual switches and alternative perceptualstates such as piecemeal rivalry and binocular fusion

Conclusions

We report that induced perceptual reversals ofillusory rotation in SFM displays are very brief andoccur in the absence of visual awareness

Keywords structure-from-motion perceptual alterna-tions multistable perception kinetic-depth effect per-ceptual reversals visual awareness

Acknowledgments

Commercial relationships noneCorresponding author Alexander PastukhovEmail pastukhovalexandergmailcomAddress Department of General Psychology andMethodology Bamberg Germany

References

Adams P A (1954) The effect of past experience onthe perspective reversal of a tridimensional figureAmerican Journal of Psychology 67 708ndash710 doi1023071418496

Alais D Cass J OrsquoShea R P amp Blake R (2010)Visual sensitivity underlying changes in visualconsciousness Current Biology 20 1362ndash1367 doi101016jcub201006015

Blake R amp Logothetis N K (2002) Visualcompetition Nature Reviews Neuroscience 3 13ndash21 doi101038nrn701

Blake R OrsquoShea R P amp Mueller T J (1992)Spatial zones of binocular rivalry in central andperipheral vision Visual Neuroscience 8 469ndash478doi101017S0952523800004971

Blake R Westendorf D amp Fox R (1990) Temporalperturbations of binocular rivalry Perception ampPsychophysics 48 593ndash602 doi103758BF03211605

Brainard D H (1997) The Psychophysics ToolboxSpatial Vision 10 433ndash436 doi101163156856897X00357

Brascamp J W amp Blake R (2012) Inattentionabolishes binocular rivalry Perceptual evidencePsychological Science 23 1159ndash1167 doi1011770956797612440100

Brascamp J W Blake R amp Knapen T (2015)Negligible fronto-parietal BOLD activity accom-panying unreportable switches in bistable percep-tion Nature Neuroscience 18 1672ndash1678 doi101038nn4130

Brascamp J W van Ee R Noest A J Jacobs RH A H amp van den Berg A V (2006) The timecourse of binocular rivalry reveals a fundamentalrole of noise Journal of Vision 6(11)8 1244ndash1256doi1011676118 [PubMed] [Article]

Brouwer G J amp van Ee R (2006) Endogenousinfluences on perceptual bistability depend onexogenous stimulus characteristics Vision Research46 3393ndash3402 doi101016jvisres200603016

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 14

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

Chong S C Tadin D amp Blake R (2005)Endogenous attention prolongs dominance dura-tions in binocular rivalry Journal of Vision 5(11)61004ndash1012 doi1011675116 [PubMed] [Article]

Enns J T amp Di Lollo V (2000) Whatrsquos new in visualmasking Trends in Cognitive Sciences 4 345ndash352doi101016S1364-6613(00)01520-5

Faisal A A Selen L P J amp Wolpert D M (2008)Noise in the nervous system Nature ReviewsNeuroscience 9 292ndash303 doi101038nrn2258

Frassle S Sommer J Jansen A Naber M ampEinhauser W (2014) Binocular rivalry Frontalactivity relates to introspection and action but notto perception Journal of Neuroscience 34 1738ndash1747 doi101523JNEUROSCI4403-132014

Gregory R L (2009) Seeing through illusions (1st ed)New York Oxford University Press

Haynes J-D Deichmann R amp Rees G (2005) Eye-specific effects of binocular rivalry in the humanlateral geniculate nucleus Nature 438 496ndash499doi101038nature04169

Kang M-S (2009) Size matters A study of binocularrivalry dynamics Journal of Vision 9(1)17 1ndash11doi1011679117 [PubMed] [Article]

Kang M-S amp Blake R (2010) What causesalternations in dominance during binocular rivalryAttention Perception amp Psychophysics 72 179ndash186doi103758APP721179

Kim Y-J Grabowecky M amp Suzuki S (2006)Stochastic resonance in binocular rivalry VisionResearch 46 392ndash406 doi101016jvisres200508009

Knapen T Brascamp J Pearson J van Ee R ampBlake R (2011) The role of frontal and parietalbrain areas in bistable perception Journal ofNeuroscience 31 10293ndash10301 doi101523JNEUROSCI1727-112011

Kornmeier J Ehm W Bigalke H amp Bach M(2007) Discontinuous presentation of ambiguousfigures How interstimulus-interval durations affectreversal dynamics and ERPs Psychophysiology 44552ndash560 doi101111j1469-8986200700525x

Laing C R amp Chow C C (2002) A spiking neuronmodel for binocular rivalry Journal of Computa-tional Neuroscience 12 39ndash53 doi101023A1014942129705

Leopold D A amp Logothetis N K (1999) Multi-stable phenomena Changing views in perceptionTrends in Cognitive Sciences 3 254ndash264 doi101016S1364-6613(99)01332-7

Leopold D A Wilke M Maier A amp Logothetis NK (2002) Stable perception of visually ambiguous

patterns Nature Neuroscience 5 605ndash609 doi101038nn851

Levelt W J (1965) On binocular rivalry Soesterbergthe Netherlands Institute for Perception RVO-TNO

Libet B (1999) How does conscious experience ariseThe neural time factor Brain Research Bulletin 50339ndash340 doi101016S0361-9230(99)00143-4

Libet B Gleason C A Wright E W amp Pearl DK (1983) Time of conscious intention to act inrelation to onset of cerebral activity (readiness-potential) The unconscious initiation of a freelyvoluntary act Brain A Journal of Neurology 106623ndash642 doi101093brain1063623

Martinez-Conde S Macknik S L amp Hubel D H(2004) The role of fixational eye movements invisual perception Nature Reviews Neuroscience 5229ndash240 doi101038nrn1348

Metzger W (2009) Laws of seeing Cambridge MAMIT Press

Mitchell J F Stoner G R amp Reynolds J H (2004)Object-based attention determines dominance inbinocular rivalry Nature 429 410ndash413 doi101038nature02584

Moreno-Bote R Knill D C amp Pouget A (2011)Bayesian sampling in visual perception Proceed-ings of the National Academy of Sciences USA 1081ndash6 doi101073pnas1101430108

Moreno-Bote R Rinzel J amp Rubin N (2007)Noise-induced alternations in an attractor networkmodel of perceptual bistability Journal of Neuro-physiology 98 1125ndash1139 doi101152jn001162007

Naber M Frassle S amp Einhauser W (2011)Perceptual rivalry Reflexes reveal the gradualnature of visual awareness PLoS One 6 e20910doi101371journalpone0020910

Nawrot M amp Blake R (1989) Neural integration ofinformation specifying structure from stereopsisand motion Science 244 716ndash718 doi101126science2717948

Noest A J van Ee R Nijs M M amp van Wezel RJ A (2007) Percept-choice sequences driven byinterrupted ambiguous stimuli A low-level neuralmodel Journal of Vision 7(8)10 1ndash14 doi1011677810 [PubMed] [Article]

Orbach J Ehrlich D amp Heath H A (1963)Reversibility of the Necker cube I An examinationof the concept of lsquolsquosatiation of orientationrsquorsquoPerceptual and Motor Skills 17 439ndash458 doi102466pms1963172439

Orbach J Ehrlich D amp Vainstein E (1963)

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 15

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016

Reversibility of the Necker cube III Effects ofinterpolation on reversal rate of the cube presentedrepetitively Perceptual and Motor Skills 17 571ndash582 doi102466pms1963172571

OrsquoShea R P Sims A J H amp Govan D G (1997)The effect of spatial frequency and field size on thespread of exclusive visibility in binocular rivalryVision Research 37 175ndash183 doi101016S0042-6989(96)00113-7

Pastukhov A (2015) Perception and the strongestsensory memory trace of multi-stable displays bothform shortly after the stimulus onset AttentionPerception amp Psychophysics 1ndash11 doi103758s13414-015-1004-4

Pastukhov A amp Braun J (2007) Perceptual reversalsneed no prompting by attention Journal of Vision7(10)5 1ndash17 doi1011677105 [PubMed][Article]

Pastukhov A amp Braun J (2011) Cumulative historyquantifies the role of neural adaptation in multi-stable perception Journal of Vision 11(10)12 1ndash10 doi101167111012 [PubMed] [Article]

Pastukhov A amp Braun J (2013) Structure-from-motion Dissociating perception neural persistenceand sensory memory of illusory depth and illusoryrotation Attention Perception amp Psychophysics 75322ndash340 doi103758s13414-012-0390-0

Pastukhov A Garcıa-Rodrıguez P E Haenicke JGuillamon A Deco G amp Braun J (2013)Multi-stable perception balances stability andsensitivity Frontiers in Computational Neurosci-ence 7 17 doi103389fncom201300017

Pastukhov A Vivian-Griffiths S amp Braun J (2015)Transformation priming helps to disambiguatesudden changes of sensory inputs Vision Research116 36ndash44 doi101016jvisres201509005

Pastukhov A Vonau V amp Braun J (2012)Believable change Bistable reversals are governedby physical plausibility Journal of Vision 12(1)171ndash16 doi10116712117 [PubMed] [Article]

Platonov A amp Goossens J (2014) Eye dominancealternations in binocular rivalry do not requirevisual awareness Journal of Vision 14(11)2 1ndash17doi10116714112 [PubMed] [Article]

Prins N amp Kingdom F A A (2009) PalamedesMatlab routines for analyzing psychophysical dataRetrieved from httpwwwpalamedestoolboxorg

Ramachandran V S amp Anstis S M (1983)Perceptual organization in moving patterns Na-ture 304 529ndash531 doi101038304529a0

Roeber U Veser S Schroger E amp OrsquoShea R P(2011) On the role of attention in binocular rivalryElectrophysiological evidence PloS One 6(7)e22612 doi101371journalpone0022612

Shpiro A Moreno-Bote R Rubin N amp Rinzel J(2009) Balance between noise and adaptation incompetition models of perceptual bistability Jour-nal of Computational Neuroscience 27 37ndash54 doi101007s10827-008-0125-3

Stonkute S Braun J amp Pastukhov A (2012) Therole of attention in ambiguous reversals of struc-ture-from-motion PloS One 7 e37734 doi101371journalpone0037734

Treue S Andersen R A Ando H amp Hildreth E C(1995) Structure-from-motion Perceptual evidencefor surface interpolation Vision Research 35 139ndash148 doi1010160042-6989(94)E0069-W

van Dam L C J amp van Ee R (2006) Retinal imageshifts but not eye movements per se causealternations in awareness during binocular rivalryJournal of Vision 6(11)3 1172ndash1179 doi1011676113 [PubMed] [Article]

van Ee R (2009) Stochastic variations in sensoryawareness are driven by noisy neuronal adaptationEvidence from serial correlations in perceptualbistability Journal of the Optical Society of AmericaA 26 2612ndash2622 doi101364JOSAA26002612

Weilnhammer V A Ludwig K Hesselmann G ampSterzer P (2013) Frontoparietal cortex mediatesperceptual transitions in bistable perception Jour-nal of Neuroscience 33 16009ndash16015 doi101523JNEUROSCI1418-132013

Wolfe J M (1983) Influence of spatial frequencyluminance and duration on binocular rivalry andabnormal fusion of briefly presented dichopticstimuli Perception 12 447ndash456 doi101068p120447

Wolfe J M (1984) Reversing ocular dominance andsuppression in a single flash Vision Research 24471ndash478 doi1010160042-6989(84)90044-0

Zhang P Jamison K Engel S He B amp He S(2011) Binocular rivalry requires visual attentionNeuron 71 362ndash369 doi101016jneuron201105035

Journal of Vision (2016) 16(3)14 1ndash16 Pastukhov amp Klanke 16

Downloaded From httpjovarvojournalsorgpdfaccessashxurl=dataJournalsJOV934914 on 02222016