Abstract
Binocular interactions within and between chromatic channels
Department of Psychology, University of York, UK
Our understanding of how the brain combines information across the eyes has been informed by detailed computational modelling of psychophysical data. Contemporary models include processes of binocular combination and interocular suppression. However, these are based primarily on experiments using achromatic sine wave grating stimuli. Here, we investigate binocular interactions within and between the chromatic pathways responding to isoluminant red/green (L-M) and blue/yellow (S-(L+M)) signals, as well as the achromatic (luminance) pathway. In Experiment 1, we replicate a complex pattern of contrast discrimination (‘dipper’) functions for four ocular arrangements of pedestal and target stimuli, which is consistent across the achromatic and chromatic pathways. In particular we find strong dichoptic pedestal masking in all three pathways, and similar levels of summation at threshold (mean of 4.6 dB). In Experiment 2 we measure dichoptic masking within and between pathways, for mask contrasts at a fixed multiple of detection threshold. We find the strongest interocular suppression within each pathway, such that the target contrast must equal or exceed the dichoptic mask contrast to be detected. We also find evidence of weaker interocular suppression between the two chromatic pathways, and the weakest suppression between the achromatic and chromatic pathways. We consider the results in the context of computational models of binocular vision, in which masking is determined by the weight of interocular suppression and the extent to which there is mandatory summation of the signals in the two eyes.
Funding & commercial relationships: Supported by BBSRC grant BB/V007580/1.
Accounting for the structure of noise in binocular rivalry
Department of Psychology, University of Warwick, UK
Binocular rivalry occurs when two eyes are presented with incompatible images; observers experience temporal alternations of the two eyes’ images. Such descriptions often mislead readers to assume regular rhythmic rivalry dynamics. Indeed, leading models of binocular rivalry, oscillator models, predict very regular dynamics. However, rivalry dynamics is noisy: percept durations show a high variability. (1) Oscillator models hardly focus on this and do not capture gamma-distributed percept durations. (2) Autocorrelations in the percept durations, although widely known, receive little attention in modelling efforts. (3) Similarly, there is a substantial chance of and individual differences on returning to the same exclusive percept after a mixture percept. These observations are unlikely to be explained by injecting Gaussian noise in oscillator models. We propose a Bayesian sampling model where percepts represent samples from an internal parameter landscape. Here, samples are autocorrelated, meaning new samples are related to preceding samples, and have momentum, meaning new samples continue in the same direction as previous samples. This captures (1) the gamma-distributed percept durations via the sampling momentum and an exponentially distributed noise; (2) the autocorrelations in percepts via autocorrelations in samples; and (3) the return rate via the symmetric nature of the parameter landscape. The Bayesian sampling model qualitatively captures the binocular rivalry dynamics that previous models did not and cannot. Binocular rivalry can be viewed as a decision-making process performed by the early visual cortex. Using Bayesian sampling models to describe such processes helps draw connections between early visual processes and high-level decision-making processes.
Funding & commercial relationships: This work is supported by a European Research Council consolidator grant (No. 817492-SAMPLING).
Computational simulations of vernier thresholds: role of fixational eye movements in visual sampling
1Department of Experimental Psychology, University of Oxford, UK
2Department of Psychology, University of Pennsylvania, USA
Fixational eye movements (FEMs) may influence visual performance through their modulation of information sampling at the earliest stage of processing. By translating the image across the retina, FEMs introduce time-varying signals and may optimise the position of the fovea with respect to the stimulus. To explore the role of FEMs we simulate the performance of a computational observer in a Vernier acuity task, judging the direction of offset between two lines. The observer has access to the mosaic of cone excitations and the eye movement path. Using a simplified, one-dimensional Vernier task, we show that motion is beneficial when the stimulus information available from the initial position is low, but detrimental when the information from the initial position is high. Next, we used ISETBio (isetbio.org) to compute Vernier thresholds in a more complete setup that simulated the two-dimensional image, optical aberrations and realistic photon capture by the cones. When the eye path was known and used ideally, we did not always observe improved performance by introducing fixational drift, and on average there was no benefit compared to the stationary condition. The lack of improvement may reflect the high-density of foveal cones relative to the optical point spread function and the fact that we did not introduce temporal filtering by the visual system. The contrast between our results and previous computational findings showing a benefit of FEMs (Anderson et al., 2020 Journal of Vision 20(7):34 1–19) suggests a rich interaction between optics, cone sampling, FEMs, post-receptoral filtering, task and visual performance.
Funding & commercial relationships: This work received a UKRI's Physics of Life grant funded by the Engineering and Physical Sciences Research Council and the Wellcome Trust [grant code: EP/W023873/1].
What makes visual search inefficient?
1University of Leicester, UK
2University of Aberdeen, UK
3University of Essex, UK
Previous research has explored whether and when eye movement choices reflect efficient resource allocation, in the sense that the eyes are directed to locations where high spatial resolution is required. One approach to answering this question is to divide the screen into two halves such that the target on one (‘easy’) side can be detected using peripheral vision, and on the other (‘hard’) side, requires central vision. Under such conditions eye movements should be directed to the hard side. Our previous research consistently found large individual differences in this measure of efficiency: some people are optimal, some stochastic and some counter-optimal. Here we explored these individual differences in a large sample (n=267) who completed the split-half search task alongside a battery of other task and measures. Many potential explanations for individual differences can be ruled out by these results, which we will discuss. We also asked participants to describe their search strategy. Over half the participants spontaneously reported using the division of the search area into two halves as part of their strategy, and most of these participants describe a deliberate strategy of looking at the easy side and then, if the target is not there, looking at the hard side. That is, participants describe using exactly the opposite of an optimal strategy. In contrast, efficient searchers are more likely to report using no strategy at all. The implication is that inefficient searchers, rather than being less thoughtful or deliberate with their eye movements, may instead be ‘over-thinking it’.
Funding & commercial relationships: Economic and Social Research Council [ES/S016120/1 to Amelia Hunt, PI, Alasdair D.F. Clarke, Co-I]
Photoreceptor mechanisms underlying human non-visual responses to light
Technical University Munich, Germany
Beyond vision, light exposure profoundly affects physiology and behaviour by synchronising our circadian clock to the light-dark cycle and suppressing the production of the hormone melatonin. These ‘non-visual’ functions are mediated by a subset of retinal ganglion cells expressing the photopigment melanopsin. Many details on retinal mechanisms mediating the non-visual effects of light remain unclear. In this talk, I will describe my research characterising the retinal mechanisms underlying ‘non-visual’ photoreception in humans.
Evidence of a third functional visual pathway and robust individual differences
School of Psychology, Cardiff University, Wales, UK
A number of groups have independently proposed that the human brain contains not two but three visual pathways (ventral, dorsal and lateral). Standard resting state analyses of BOLD data, which fractionate the brain into neural networks, isolate dorsal and ventral visual networks, but not a separate lateral one. We investigated whether a third functional visual network would be revealed during viewing of naturalistic visual stimuli (a Hollywood movie). A lateral network component was indeed found to segment off from a standard dorsal network component when group-level ICA (Independent Component Analysis) was performed on the movie data. Interestingly when individual-level, rather than group-level, ICAs were performed, the third pathway was only found in approximately 2/3rds of brains, it was absent or indistinct in the remainder. We confirmed the robustness of our findings with additional analyses, and a further independent dataset, HBN-SSI. In addition to movie data, the HBN-SSI dataset includes data from observers watching abstract geometric animations (‘Inscapes’). Group-level analysis revealed a lateral component in the movie data but not in the Inscapes data. These results provide compelling evidence of a third functional pathway that is activated during viewing of naturalistic stimuli. The finding of individual differences is intriguing and raises the question of whether apparent differences in functional architecture might have an impact on perception and behaviour.
Funding & commercial relationships: ESRC grant (ES/S015272/1)
Information loss and accuracy gain in redundancy masking
1CNRS, University of Lille, SCALab, UMR9193, Lille, France
2Department of Psychology, Sabanci University, Istanbul, Turkey
Redundancy masking (RM) is the compression of visual information in repeating patterns: the perceived number of items is smaller than the presented number of items, even when only three items are presented. For example, observers often report only two bars when presented with three identical bars in the visual periphery. Here, we investigated to what extent features of masked items survived RM or were lost entirely. We presented three to five radially arranged bars with varying widths for 150 ms to the left or right of fixation. Observers first reported the perceived number of bars, and second, adjusted the widths of – and spacings between – probes to match what they perceived. We computed deviation scores for the reported number of bars, widths and spacings between the bars. Deviation scores for the number of bars showed strong RM. Overall, the width of thin (thick) bars tended to be overestimated (underestimated). In RM trials, the reported bar width was slightly larger than in correct (non-RM) trials. Interestingly, the reported width of thicker bars was more accurate in RM trials than in correct trials. The reported spacing between lines was larger in RM compared to correct trials (especially for three presented lines), replicating previous results that showed compressions of visual space in RM. Our results suggest that erroneous perception of lower numbers of items in RM may go hand in hand with higher accuracy in reporting their features. We discuss our results in light of potential benefits of RM beyond the economical use of limited processing capacities.
Funding & commercial relationships: ANR-19-FRAL-0004; TUBITAK122N748
Exploring the motion deblurring mechanism from a visual prediction perspective
School of Psychology, University of Nottingham, UK
Objects in motion are subject to ‘visual persistence’ causing motion blur (Burr, 1980 Nature 284 164–165). It has been suggested that the visual system might aggregate and summate an object's spatial pattern along the motion path to mitigate motion blur. However, the exact mechanism underpinning this proposal remains unknown. One possibility is that space-time summation is implemented passively by space-time-oriented filters in the visual cortex (Nishida et al., 2007 Curr Biol 17 366–372). An alternative possibility is, from the visual prediction perspective, that the visual system may actively use motion signals and the prediction mechanism to shift forward spatial pattern, subsequently averaging aligned signals to implement deblurring. We compared the apparent contrast of a drifting random noise pattern and a randomly updating noise pattern that were indistinguishable in static frames and in terms of retinal dynamics. We found an increase in apparent contrast for drifting versus randomly updating noise. The contrast enhancement effect depended on image speed, which undermines the passive matched filter account, assuming a separate channel for each speed. Furthermore, using the same stimuli, in which the translating region was sandwiched between updating regions, we observed a motion-induced shift for translation, the size of which was correlated, over subjects and speed conditions, with the contrast enhancement effect, indicating a firm linkage between the two phenomena. Overall, these findings show an unexpected connection between motion-induced spatial shifts and the contrast enhancement effect, suggesting the participation of visual prediction in preventing motion blur.
Funding & commercial relationships: Dan Hu was supported by a University of Nottingham, Vice Chancellor's scholarship for Research Excellence (international).
Pointillistic perception: the local nature of suprathreshold visual processing
Division of Optometry and Vision Sciences, School of Health Sciences, City University of London, UK
Kapadia, Westheimer & Gilbert (2000, J Neurophysiol) reported that macaque cortical receptive fields contract towards a local punctate form for high contrast stimulation. To study this reorganization in human, we used Gabor arc at uniform eccenticity of 3 deg with half-height of 0.28 deg, carrier spatial frequency of 2.5 cy/deg and variable length from 0.05 to 5 deg. Detection thresholds for these elongated Gabor stimuli were compared on a blank background and on a suprathreshold 50% contrast uniform Gabor ring of the same spatial frequency as a function of the length of the Gabor arc. Summation was linear at threshold for arc lengths up to about 6 arcmin, then followed a square-root law up to about 2 deg, implying the availability of summation mechanisms (=receptive fields) within that range, with no further summation for longer arcs. On the suprathreshold background, there was no significant summation for a substantial distance beyond the short linear range, implying that only the smallest receptive fields were available to perception up to about the 2 deg length. Beyond this point, however, a second linear summation mechanism came into play, extending to about 4 deg length, suggesting that an extended range of length-summation mechanisms, or receptive fields, is available at contrast detection threshold, but in everyday suprathreshold conditions the human spatial processing system is limited to just a local punctate summation field and a 40:1 elongated summation field. This organization can account for the evanescent curved lines of the recently-popular Porche illusion.
The art and science of Patrick Hughes’ ‘Reverspectives’
1Experimental Psychology, University of Oxford, UK
2Reverspective, London, UK
In Patrick Hughes’ Reverspectives, the disparities created by the protruding, truncated pyramids contradict the perspective provided by the images on the sides of the pyramids. Perspective typically dominates so that observers perceive the scene to be receding into the distance when Reverspectives are viewed from >1 m. To investigate this effect systematically, we created virtual Reverspectives in which the disparity and perspective information could be independently manipulated, as well as the motion parallax created when the observer moves from side-to-side. In a virtual Reverspective, there is no ‘real’ depth. The first experiment investigated the relative strength of perspective and disparity information. To do this, we created virtual Reverspectives that contained an adjustable blend of (i) the receding perspective scene and (ii) a random dot array depicting a protruding 3-D structure. The threshold for perceiving the disparity-specified structure typically occurred when the blend of perspective scene to random dots was around 50%:50%. In a second experiment, observers were asked to report whether the 3-D structure appeared to recede or protrude when the blend was 30%:70% (i.e., overwhelmingly random dots). Not surprisingly, most observers reported that the 3-D structure appeared to protrude. However, when observers were asked to move from side-to-side, all observers reported that the perceived 3-D structure reversed in depth and rotated so as to follow the observer's side-to-side movements. Hence, the strength of the perspective determines the way motion parallax is interpreted, despite the initial appearance of a protruding structure.
Perception of self-motion in microgravity
1Centre for Vision Research, York University, Toronto, Canada
2Hochschule Bonn-Rhein-Sieg, Germany
3Justus Liebig University Giessen, Germany
When perception of self-motion is induced using only visual motion, vestibular cues indicate that the body remains stationary which may bias an observer's perception. When lowering the reliability of the vestibular cue by for example lying down or by adapting to microgravity, these biases may decrease, accompanied by a decrease in precision. To test this hypothesis, we used a move-to- target task in virtual reality. Astronauts (6 female, 6 male) and Earth-based controls (10 female, 10 male) were shown a simulated target at a range of distances. After the target disappeared, self- motion was induced by optic flow. Participants indicated when they thought they had arrived at the target's previously seen location. Astronauts completed the task on Earth (lying supine and sitting upright) prior to space travel, twice in space onboard the International Space Station (ISS) (within 3–6 days of arrival, after ∼85 days in space), and after landing (within 3–6 days of return and ∼85 days later). Controls completed the experiment on Earth using a similar regime with a supine posture used to simulate being in space. While variability was similar across all conditions, astronauts displayed higher gains (target distance/perceived travel distance) when lying supine than when sitting upright in two out of the three terrestrial sessions. No difference could be detected between the astronauts’ performance on Earth and onboard the ISS or between the controls’ sessions. Despite adapting to a floating mode of travel in the ISS, astronauts’ performance in judging self-motion distance appears largely unaffected by exposure to microgravity.
Funding & commercial relationships: Canadian Space Agency (15ILSRA1-York) Natural Sciences and Engineering Research Council of Canada.
The relationship between serial dependence and individual differences in multiple object tracking
1School of Psychology, University of Lincoln, UK
2School of Psychology, University of Aberdeen, King’s College, Aberdeen, UK
3Lincoln Medical School, University of Lincoln, UK
Serial dependence represents a perceptual process wherein prior perceptual information influences subsequent perceptions, facilitating a stable interpretation of our environment. The underlying mechanisms of serial dependence have been heavily debated, from mere attentional resources to decision-making task-related processes. Here, we use an independent measure of attention, the Multiple Object Tracking (MOT) task, to investigate whether tracking can predict the serial dependence in position perception, that is the bias towards the previously presented stimulus locations. In the MOT task, 50 participants were asked to track multiple moving objects on the screen and ignore distractors. In the position adjustment task, a grating was displayed at a random location, and participants reported its perceived location. We then analysed each participant's tracking performance and the ability to report the grating position. Our findings indicate that tracking MOT performance was positively correlated to performance in the position adjustment task. To further explore individual differences, we investigated serial dependence strength in individuals with high versus low MOT performance, such as video-game players and non-gamers. Our results indicate that tracking abilities were positively correlated with serial dependence strength in the gamer group.
Eye movements, pupil size and scene perception in real-world indoor and outdoor scenes
1Max Planck Institute for Biological Cybernetics, Germany
2Eberhard Karls University of Tübingen, Germany
3Technical University Munich, Germany
While eye movement behaviour in well-constrained laboratory studies is very well understood, we know much less about eye movements under naturalistic conditions. Here, we created a benchmark dataset for naturalistic free-viewing, aiming to investigate differences in saccade, fixation and pupil size metrics between indoor and outdoor environments. Using a head-mounted wearable eye tracker (Tobii Glasses Pro 3), participants engaged in a naturalistic free-viewing task (Kay et al., 2023).They viewed six diverse indoor and outdoor scenes while withholding information about the nature of the experiment from the participants (n=18, 13 female; mean age 24.2 years). For each scene, we recorded eye movements during a 4-min measurement interval while the light metrics of the scenes were concurrently recorded using a spectroradiometer. After each measurement, we also asked participants to rate the scene in terms of its beauty, brightness and complexity, and rate their sleepiness. Light levels of indoor scenes were significantly lower (520.14±351.16 photopic lux) than those of outdoor scenes (9437.56±8822.15). Pupil size was substantially larger in indoor (2.88±0.71 mm) compared to outdoor conditions (2.07±0.32 mm). We found no significant differences in any of the eye movement metrics. In an exploratory analysis, we found a complex landscape of correlations between various eye movement, pupil and light metrics. We found no significant differences in naturalistic free-viewing of quantitatively and qualitatively different outdoor and indoor scenes. Our data set can serve as a benchmark for future eye movement studies.
What are the effects of constant and varying visual feedback delays on a simple target acquisition task?
1School of Psychological Science, University of Bristol, UK
2Department of Psychology, University of Bath, UK
When using interactive technologies, data transfer and processing times cause delays between our actions and the system response. These delays can be constant but can also vary moment-by-moment. Here, we report two studies that investigate the effect of these two types of delays on a visually guided task. We recorded the participant's cursor movements and eye movements as they were exposed to delayed visual feedback while completing a target acquisition task. In Experiment 1, there was a constant visual feedback delay. The participants initially overshot the target but generated increasingly accurate movements with repeated exposure. Following the removal of the delay, aftereffects were observed – the participants now systematically undershot the target. This is a signature of delay adaptation in the closed-loop control system. The participants also demonstrated a tendency to rapidly fixate each target throughout the exposure phase, suggesting eye movements were not important for delay adaptation. In Experiment 2, there was a varying visual feedback delay. Across the two experiments, the participants showed the same pattern of adaptation and gaze behaviour. Despite this, the varying delay group demonstrated smaller post-exposure impairments in task completion time and movement velocity. We show that strict control of the movement speed is not necessary for delay adaptation. We also provide evidence that adaptation to varying delays is possible and that the fluctuation did not affect the adaptation rate. These findings suggest that delay adaptation is robust across a wider range of tasks and conditions than previously considered.
Funding & commercial relationships: BT Group plc & the EPSRC
Measuring the spatiotemporal contrast sensitivity function in infantile nystagmus
School of Optometry and Vision Sciences, Cardiff University, UK
Infantile nystagmus (IN) is characterised by a continuous, involuntary oscillation of the eyes and is commonly associated with visual impairment. Interventions which reduce the intensity of IN oscillations can lead to a subjective improvement in vision, but this is not always accompanied by a corresponding improvement in the conventional clinical measure of visual acuity. However, visual acuity gives a limited picture of visual function, concerning only high spatial frequencies which are stationary and viewed at full contrast. A more comprehensive overview of visual function may be provided by the spatiotemporal contrast sensitivity function (stCSF), which describes the limits of vision in terms of contrast, velocity and spatial frequency. Here, we present a novel method to measure the stCSF by measuring contrast sensitivity functions (CSFs) at different velocities using the optokinetic response (OKR). We firstly demonstrate the feasibility of using this method to measure the stCSF in control participants by reproducing established effects of velocity on the CSF. Secondly, we compare CSFs measured at midrange velocities between 14 participants with IN and 15 age-matched controls, revealing a shift of the CSF in IN compared to controls. This may indicate impaired contrast sensitivity in IN, possibly as a consequence of their involuntary eye movements. Our method, adapted to utilise the reflexive OKR, could be used in paediatric and adult populations to better assess the efficacy of therapeutic interventions for IN.
Funding & commercial relationships: College of Optometrists Postgraduate Scholarship
The impact of multisensory and uni-sensory cue placement in a multiple object tracking task
1University of Lincoln, School of Psychology, UK
2University of Lincoln, School of Computer Science, UK
3University of Staffordshire, School of Computer Science, UK
Previous research has established that the addition of sensory cues placed on targets in a multiple object tracking (MOT) task, allow adults to track a higher amount of moving targets amongst indistinguishable distractors, compared to when no sensory cues are presented (Föcker et al. 2022, Attention, Perception and Psychophysics 84 (5) 1611–1624). The aim was to investigate whether sensory cues can guide attention in a bottom-up or top-down mechanism during tracking. Sensory cues were presented on targets or on distractors. We hypothesised that when the sensory cues were placed on distractors it would impair tracking performance as attention would be automatically guided towards the elicited sensory cues. Thirty-three participants were asked to keep track of targets amongst indistinguishable distractors. All objects bounced against a central circle and elicited visual, auditory and audio-visual cues or no cues. The results showed that visual cues presented on targets improved attentional tracking capacity, whereas tracking performance was impaired when the visual cues were presented on the distractors. As tracking performance did not differ when visual cues were presented on the distractors compared to visual cues absent, performance was modulated by attentional load (Lavie, 1995, Journal of Experimental Psychology: Human Perception and Performance, 21(3) 451–469). Tracking performance improved when auditory cues were absent compared to present. To conclude, visual cues might guide attention in a top-down mechanism when placed on targets but this mechanism is impaired when placed on distractors due to bottom-up mechanisms of alerting the attention away from the targets and attentional load.
A magno-biased adaptable metric mediates perception of spatial separation
1School of Psychology, University of Nottingham, UK
2Laboratoire des Systèmes Perceptifs, Ecole Normale Supérieure, PSL University & CNRS, Paris, France
Adapting to a dynamic pattern (e.g., of flickering dots) can induce a compression in the perceived separation between subsequently presented dot pairs (Hisakata et al., 2016 Current Biology 26(14) 1911–1915) implying the early modification of an internal spatial metric in the visual system (Jovanovic et al., 2022 Journal of Vision 22(11):7 1–23). Here, we manipulated the content of the adapting pattern in terms of its spatial frequency, temporal frequency, colour and motion to identify the origin of the compression after-effect. Participants initially viewed an adapting pattern of dots presented in one hemifield, then compared the separation of a pair of dots presented at the adapted location and a pair presented at the same eccentricity in the unadapted hemifield. We found that adaptation to dots that were flickering in luminance at high-temporal frequency (60 Hz), and were consequently invisible, induce a spatial compression. For Gabor-array adaptors, compression is strongest for lower spatial frequency carriers. Isoluminant colour adaptors induce compression at low, but not high, temporal frequency. When moving dots are used to induce the compression after-effect, a reduced magnitude of compression coincides with a breakdown of perceived motion coherence at isoluminance. Together, these results indicate that the observed adaptation operates within spatially low-pass, temporally band-pass transient luminance-based channels, characteristic of the preferred sensitivities of the magnocellular pathway. We infer that neural processes responsible for encoding metric properties of spatial vision can be influenced by adaptive mechanisms in the magnocellular pathway.
Funding & commercial relationships: Leverhulme Trust
From semantics to saccades: analysing the impact of semantic contrast on eye movements
1School of Biological and Behavioural Sciences, Queen Mary University of London, London, UK
2Department of Psychology, Northeastern University, Boston, USA
3LIRIS, CNRS, University of Lyon, France
The acquisition of visual information is typically continuous over time, allowing for the accumulation of knowledge about unfolding events. When this continuity is disrupted, observers exhibit changes in their eye movement patterns. However, the factors influencing these differences remain largely unexplored. We propose that such differences may be driven by the extent to which the semantic content of the visual input changes. As such, we used computational linguistics (GloVe; Pennington et al., 2014), to develop a measure to quantify how semantically different two images are, which we term ‘semantic contrast’. We applied this metric to analyse data from an eye tracking experiment in which 48 participants freely viewed 80 sequences of film frames. Each sequence contained several frames depicting a consistent unfolding of events, followed by a frame that was either a natural continuation of the preceding ones (Continuous condition) or was unrelated to them (Discontinuous condition). Our analysis revealed that in the Discontinuous condition only, as semantic contrast increased, there was a decrease in the average number of fixations (p = .04) and interfixation distance (p = .001), while the average fixation duration increased (p = .007). These altered eye movement patterns suggest heightened attentional focus, potentially driven by the additional processing challenge associated with large changes in semantic content. As a result, our study shows the utility of our approach in capturing the degree of semantic content change between images and demonstrates the relationship between semantic contrast and eye movements.
Funding & commercial relationships: The research is funded by a Leverhulme Trust grant (RPG-2020-024) awarded to Isabelle Mareschal, Peter Bex and Antoine Coutrot.
The neurochemistry and connectivity in the brain in charles bonnet syndrome
1Wellcome Centre for Integrative Neuroimaging, Nuffield Department of Clinical Neurosciences, Oxford, UK
2Vision and Eye Research Institute, Anglia Ruskin University, Cambridge, UK
Charles Bonnet Syndrome (CBS) is a condition in which people with vision loss experience complex visual hallucinations. Our study aimed to investigate the neural basis of these hallucinations. Specifically, we investigated the association between γ-aminobutyric acid (GABA) in the early and ventral visual cortex and the presence of visual hallucinations in CBS to evaluate if cortical disinhibition plays a role in the condition. We applied multimodal 3T MRI, including task and resting-state functional MRI in 12 participants with vision loss, 6 with and 6 without CBS. GABA was measured in two locations, the early visual cortex and the lateral visual cortex. All participants additionally completed visual acuity tests, cognitive tests, and a visual hallucination questionnaire. There was no difference in the concentration of GABA between the participants with and without CBS in either the early visual or the lateral visual cortex. All participants showed BOLD activation to images of objects and the phase-scrambled control. Analysis of resting-state functional connectivity with lateral visual cortex and early visual cortex showed little difference between the CBS and non-CBS participants. A less stringent fixed-effects analysis showed a correlation between the extent of functional connectivity with LOC and hallucination strength. Overall, our results provide no strong evidence for an association between visual cortex GABA and CBS and any differences in visual function and connectivity were not sufficient to emerge in the small population studied.
Funding & commercial relationships: This research was supported by The Royal Society (Dorothy Hodgkin Research Fellowship to IBI) and the Medical Research Council (MR/V034723/1), Fight For Sight [24BV20], Esme's Umbrella, Blind Veterans, The Royal Society, National Institute of Health Research (via portfolio funding), and College of Optometrists and supported by the NIHR Oxford Health Biomedical Research Centre (NIHR203316). The Wellcome Centre for Integrative Neuroimaging is supported by core funding from the Wellcome Trust (203139/Z/16/Z and 203139/A/16/Z).
Spatial summation within static and dynamic glass patterns
1Eye and Vision Research Group, University of Plymouth, UK
2Centre for Cognitive Neuroscience, College of Health, Medicine and Life Sciences, Brunel University London, UK
Neurons in early visual cortex are tuned to local stimulus features, that is, orientation and spatial frequency. Behaviourally, however, the stimuli that are ecologically relevant are more complex. The visual system must therefore employ mechanisms that combine these localised outputs. Previous studies addressed this question by measuring area summation for various types of Glass pattern, in which the stimulus array was divided into various sectors of increasing size. Linear summation predicts that the decrease in thresholds with increasing signal area follows a power-law function with a (log-log) slope of −1.0. This was previously found for circular Glass patterns, suggesting that the information in these patterns is linearly summed across the entire display, and was taken as evidence for specialized global detectors for circular patterns. Here we present results from summation experiments using static and dynamic Glass patterns with radial, circular and translational structure. We measured detection thresholds as a function of signal area (25%, 50%, 100%) with the remaining area containing orientation noise. Results show that thresholds for dynamic Glass patterns were lower than for static ones across all pattern types and signal areas. In agreement with Schmidtmann et al. 2015 J Vis 15(16):6 1–19, we do not find superior sensitivities for detecting circular structure. Importantly, summation strength never reached values close to linear summation (−1) for static or dynamic patterns, implying that probability summation is the most likely basis for the detection of circular, radial and translational Glass patterns.
What does the visual brain see after loss of photoreceptors?
Department of Biomedical Engineering, Robert R. McCormick School of Engineering and Applied Science, Northwestern University, USA
It has been known for some time from animal models that retinal circuits undergo rewiring following the loss of photoreceptors. There is also growing evidence that these rewired circuits can produce firing patterns in retinal ganglion cells that one supposes would be interpreted by higher visual centres as signals that might evoke perceptions. To our knowledge there are no reports of such perceptions in humans with photoreceptor loss but that might be simply because no-one has considered what perceptions might be expected based on animal studies. We have studied retinal ganglion cell discharges in the S334ter rat, a transgenic strain with profound photoreceptor loss. The retinal ganglion cells in these animals display rhythmic firing that is correlated for some distance over the retina. Both positively and negatively correlated activity is observed, suggesting that OFF and ON retinal circuits may be engaged differentially, just as they are in the normal retina. Based on these results one might expect perceptions of local flicker at the rhythmic frequency to occur. Regardless of the presence or absence of flickering visual perceptions, the unstimulated rhythmic discharge of retinal ganglion cells presents a challenge that should be considered in the design of retinal prostheses.
Funding & commercial relationships: NIH, NSF. Christina Enroth-Cugell and David Cugell Fellowship (JY).
Confidence biases in perceptual decision-making
Royal Holloway, University of London, UK
Confidence biases, characterized by over- or under-estimation of uncertainty, are considered an important source of errors in cognitive tasks based on knowledge or experience. However, it is less clear whether systematic biases are also present in perceptual decisions. Traditional approaches to study confidence in perceptual decisions rely on self-report measures, which are limited in their ability to characterise confidence biases by the need for unverifiable assumptions about the translation of internal confidence states into explicit ratings. I will discuss an alternative approach using implicit measures of confidence inferred from computational modelling of choice behaviour. By analysing how confidence informs sequences of decisions through a dual-decision method (Lisi et al., Nature Human Behaviour 5, 273–280), we can attain a principled measure of confidence biases relative to the idealized benchmark of a Bayesian observer. Available data obtained using this approach reveals a consistent pattern of under-confidence in perceptual decision-making compared to this Bayesian standard, indicating a tendency for individuals to underutilize their sense of confidence. Moreover, I will discuss ongoing work which suggests that these under-confidence biases might be interpreted as rational adaptation by the brain, aimed at compensating for its intrinsic limitations in accurately evaluating decision reliability. This research provides insights into the mechanisms underpinning confidence biases in perception, offering implications for our understanding of how confidence biases may form and maintain.
Funding & commercial relationships: BA/Leverhulme and Royal Holloway start-up
The role of image statistics and personality traits on perceived intimacy of architectural spaces
1Department of Psychology, Heriot-Watt University, UAE
2Institute of Anatomy I, Jena University Hospital, Germany
Perceived intimacy and aesthetic judgements of architectural spaces have been shown to be influenced by perceived visual features such as contour curvilinearity, space openness, ceiling height, and window-to-wall ratio, while individual differences in personality traits and architectural expertise modulated these effects (Ho & Hayn-Leichsenring, XXVI Conference of the International Association of Empirical Aesthetics, 2021). On the other hand, certain measurements of image statistics (e.g., edge-orientation entropy) were found to account for more variances than perceived features (e.g., curvilinearity) in pleasantness judgements and preferences for a wide range of man-made images including artwork, architecture and simple patterns (Grebenkina et al., 2018 Front. Neurosci., 12, 678). Here, we asked whether objective image statistics are stronger predictors of perceived intimacy ratings of residential building exteriors and indoor spaces than perceived features and whether image statistics interacts with the observers’ attributes in similar ways as reported previously. To this end, we presented the participants (N=790) 40 images of building exteriors and 40 images of indoor spaces and asked them to rate each image on perceived intimacy. Results of linear mixed model analyses show edge density (related to perceived complexity) was the only significant predictor of the participants’ intimacy ratings. However, there are multiple interactions between the personality dimensions (e.g., ‘openness to experience’ and ‘agreeableness’) and objective image statistics (e.g., edge density, hue and anisotropy). Taken together, our findings suggest that affective judgements of the built environment are the product of the beholder's attributes as well as basic image statistics that underlie perceived visual features.
Funding & commercial relationships: This research was funded by the Institute of Anatomy I, University Hospital Jena.
The pint glass illusion: Exploring the misjudgment of curvature and length for two and three-dimensional objects
Eye and Vision Research Group, University of Plymouth, UK
Humans are exquisitely sensitive in curvature detection and discrimination, but misjudge the extent of curvature. A largely unexplored phenomenon is the underestimation of the circumference length of curved objects. Anecdotal, and some experimental evidence (Rivera-Aparicio & Firestone, 2018 J Vision 18(10) 1318–1318), suggests that when comparing the circumference of a Pint glass to its height, subjects significantly underestimate the circumference (Pint Glass Illusion, PGI). I present results from two experiments exploring the dependence of the PGI on segment length and type of curve for 2D and 3D shapes. (i) 10 subjects were presented with circular and elliptical arcs (2D) of varying segment length (central angle, CA = 45°–360°) and aspect ratios (AR = 1, 0.5, 0.25). Subjects adjusted a straight line to the arc length. Contrary to the PGI, results show that for all segment lengths, subjects on average overestimated the length with the largest overestimation for CA = 90° and 180° (∼40%), with the most accurate estimation for the full circle and ellipses (∼20%). (ii) To investigate whether results were due to the planar presentation, subjects also judged the arc length of 3D cylindroids (height 14 cm). For elliptical cylindroids with AR = 0.25 (CA = 45°, 90°), subjects underestimate the arc length (∼ 8%). For AR = 0.5 (CA = 45°and 90°), observers overestimate the arc length after which the performance is very accurate for semi and full ellipses and circles. The observed overestimation and most accurate performance for full ellipses and circles implies that the PGI might depend on different strategies, for example, the comparison of height to circumference.
Common patterns of spatial selectivity in early visual cortex and face-selective brain regions
1Experimental Psychology, University College London, UK
2Institute of Ophthalmology, University College London, UK
3School of Optometry and Vision Science, University of Auckland, New Zealand
4Psychological Sciences Research Institute, UCLouvain, Belgium
5Institute of Neuroscience, UC Louvain, Belgium
Face recognition is widely considered to be ‘special’, involving dedicated brain regions with unique patterns of selectivity. Despite this, we recently found that the resolution of face perception was better along the horizontal versus vertical meridian, and the lower versus upper visual field, mirroring spatial variations typically found for low-level vision (Morsi et al., 2023 bioRxiv). This suggests that face-selective brain regions may vary in their retinotopic sensitivity around the visual field in the same way as lower-level areas, despite their apparent uniqueness. To investigate this we carried out retinotopic mapping of early (V1-V3) and face-selective cortical regions (OFA, pFus, mFus) using large-field bars of faces (21° eccentricity), with spatial selectivity measured via population receptive field (pRF) analysis. While pRFs are considerably larger in face-selective regions than in V1-V3, their size did not consistently vary in line with the behavioural anisotropies in any cortical regions. However, both early cortex and face-selective areas show a greater number of pRFs, with a concomitant increase in visual field coverage (an estimate of visual field sampling incorporating both pRF number and size), along the horizontal versus vertical meridian and in the lower versus upper field. The variations in our face-recognition abilities around the visual field could thus be driven by this anisotropic sampling. The commonality of these variations in visual field sampling between face-selective cortex and earlier visual regions further support a hierarchical model whereby the spatial selectivity of higher-level areas is built upon the selectivity of lower regions, even for specialised face processing.
Funding & commercial relationships: Biotechnology and Biological Sciences Research Council [grant number BB/J014567/1].
Mouse tracking as a potential alternative for eye tracking in research on individual differences in image-viewing behaviour
1School of Biological and Behavioural Sciences, Queen Mary University of London, London, UK
2Department of Psychology, Northeastern University, Boston, USA
3LIRIS, CNRS, University of Lyon, France
4Experimental Psychology, Justus Liebig University, Giessen, Germany
*Equal contribution
When a group of people view an image, they usually attend to only several of its regions. This group-level consistency, however, is underpinned by inter-individual variability: individuals differ regarding the extent to which they are drawn to specific types of semantic image content (e.g., faces vs. text). We examined if these idiosyncrasies are measurable without eye tracking. Specifically, we tested if they manifest when individuals explore blurred images using a mouse-contingent aperture unblurring part of the image. Our experiment consisted of two sessions. In the first session, 60 participants free-viewed a set of unblurred images typically used for measuring individual differences of interest. A week or more later, the same individuals freely explored blurred versions of the same images using the mouse-contingent unblurring. For each exploration modality (eyes and mouse), we quantified the time each participant spent dwelling on different types of semantic content. We did it for each half of the images separately and correlated these times across participants between the halves. Within each modality, participants were consistent in their tendencies to focus on different types of content: the correlations amounted to ∼0.9 for faces and were between ∼0.6 and ∼0.8 for other content types. Next, we correlated the individual proportions of dwell times between modalities (eyes vs. mouse). Surprisingly, these correlations were much lower (between ∼0.2 and ∼0.5), indicating that the structure of the individual differences was not preserved between modalities. We hypothesise that this divergence was due to the different motor costs of eye and mouse movements.
Funding & commercial relationships: Funded by a Leverhulme Trust grant (RPG-2020-024) to IM, PB, and AC, European Research Council Starting Grant 852885 INDIVISUAL (BdH), Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) Project No. 222641018-SFB/TRR 135 TP C9 (BdH) and “The Adaptive Mind,” funded by the Excellence Program of the Hessian Ministry of Higher Education, Science, Research and Art (BdH).
Validation of a portable arduino-based device to measure the rayleigh match
Department of Experimental Psychology, University of Oxford
Commercially available Anomaloscopes are precise, well-validated tools that allow for red-green colour vision deficiency testing using the Rayleigh equation (e.g., Jägle at al., 2005 Graefes Arch. Clin. Exp. Ophthalmol. 243 26–32). The Rayleigh match can also be used to estimate photopigment spectral sensitivities for colour normal observers (Thomas & Mollon, 2004 Visual Neuroscience 21(3) 477–482). Limited research indicates the distribution of photopigment spectral sensitivities may vary between ethnic groups within the colour normal population (e.g., Hagen et al., 2019 Vision Research 162 20–28). Historically, it has been difficult to study large, diverse populations in lab settings. A more portable and cheaper Anomaloscope would be useful for field testing, with the aim to increase sample sizes and ethnic diversity. One candidate for this is a custom-built, Arduino-based Anomaloscope (Keesing & Brainard, https://github.com/BrainardLab/TeachingCode/tree/master/ArduinoAnomaloscope). We test the suitability of this device for conducting Rayleigh matches in the colour normal population. Twenty-eight participants made five Rayleigh match settings on the Arduino device and on an Oculus HMC Anomaloscope. To investigate test-retest reliability, the experiment was repeated an hour after the initial measurements and a week later. Both devices showed high inter-session correlations (all p < .001). Participants' settings on the two devices were also highly correlated (r = .83, p < .001). These results suggest that the Arduino device is a reliable tool for measuring individual differences in the Rayleigh match in the colour normal population. We discuss future plans and possibilities for using this Arduino-based device in colour vision research.
Funding & commercial relationships: UKRI Physics of Life (EPSRC and the Wellcome Trust): EP/W023873/1; The Wellcome Trust 218657/Z/19/A; Department of Experimental Psychology, University of Oxford
The influence of reachability on judged graspability and size
1Cardiff University, UK
2Vrije Universiteit Amsterdam, Netherlands
A fundamental distinction between perception within reachable and non-reachable space has been proposed. Different cortical networks are implicated (Cléry et al., 2015, Neuropsychologia, 70, 313–326) and alignment effects (lower reaction time when reaching for virtual objects aligned for easy grasp) are only found in reachable space (Costantini et al., 2010, Experimental Brain Research 207, 95–103). Our study builds on this work to investigate whether there are differences in perceived ‘graspability’ (an affordance) in reachable versus non-reachable space. Participants were asked to judge whether they could grasp (pick up cleanly with one hand) a series of spheres with differing diameters. Participants judged they could grasp larger spheres within reach space than beyond. Biases in perception of distance and size are common in the literature on visual perception. To address the possibility that differences in graspability between reachable and non-reachable space is linked to size perception rather than affordance perception, in a second experiment we asked participants to compare the size of objects within reachable space to that of objects in non-reachable space. We found participants perceived spheres as smaller when they were in a reachable location compared to a non-reachable location. We conclude that the simplest explanation for differences in perceived graspability in reachable versus unreachable space is differences in size perception. However we do acknowledge that it could be argued that the relationship is reversed (inability to reach an object due to it being outside of reach space leads to it being perceived as larger and hence less graspable).
Funding & commercial relationships: Cardiff University School of Psychology; ESRC grant (ES/S015272/1)
Cortical contrast sensitivity across the visual field to low and high spatial frequency stimuli
1Institute of Ophthalmology, University College London (UCL), London, UK
2Birkbeck/UCL Centre for NeuroImaging, London, UK
3Department of Psychology, Royal Holloway, University of London, London, UK
4Institute of Cognitive Neuroscience, UCL, London, UK
5Department of Experimental Psychology, UCL, London, UK
Recent advances in treatments for eye diseases have raised a growing interest in mapping the knock-on effects of disease and recovery on cortical function (e.g., Farahbakhsh et al., 2022 Brain 145 3803–3815). Here we develop a novel method for measuring visual sensitivity across the visual field at the individual level, using fMRI in normal-sighted individuals. Previous studies in normal sighted, mapping well-known changes in visual sensitivity across the visual field at a cortical level (e.g., Broderick et al., 2022 J Vis 22(4):3 1–21; Liu et al., 2006 J Vis 6(11) 1294–1306), have typically only measured central locations, and using maximum stimulus contrast. This limits sensitivity to small changes in cortical visual function across the wider visual field that may be most relevant in clinical scenarios. We aim to bridge these gaps by offering a novel, sensitive method for assessing cortical visual field loss ‘in one-go’ across a large expanse of the visual space, using pRF mapping combined with contrast sensitivity estimation in V1 neurons, up to 40° diameter. Our approach can recover inhomogeneities in cortical sensitivity across the visual field at the individual level, capturing (1) how spatial frequency preference changes across eccentricities, (2) how sensitivity changes along the cardinal meridians, and (3) how simulated central loss of retinal input is reflected at the cortical level. We present a novel, effective way to measure visual field sensitivity in individual participants. This can ultimately improve understanding of how eye disease and treatment affect signalling from the retina to visual cortex.
Funding & commercial relationships: Moorfields Eye Charity – PhD studentship (GR001315); Birkbeck–UCL Centre for NeuroImaging (BUCNI), London, United Kingdom
Child-friendly vision assessments using steady-state EEG and calibration-free gaze tracking
1Institute of Ophthalmology, University College London, London, UK
2Experimental Psychology, University College London, London, UK
3NIHR Moorfields Biomedical Research Centre, London, UK
4The Francis Crick Institute, London, UK
5Department of Psychology, University of Amsterdam, Amsterdam, Netherlands
Recent breakthroughs in gene therapy have demonstrated immense promise for treating Inherited Retinal Diseases (IRDs), one of the most common causes of blindness worldwide. These disorders affect the structure and function of the retina and often lead to complete loss of light sensitivity early in life. Despite therapeutic advances, there remains a gap in efficiently assessing treatment efficacy, primarily because we lack objective and reliable vision assessments for very young patients. To address this challenge, we developed an innovative child-friendly approach combining electroencephalography (EEG) and deep-learning-driven gaze tracking to measure visual acuity in a non- invasive, reliable, and objective manner. Our protocol involves recording steady-state visual evoked potentials (ssVEP) to flickering sinusoidal gratings embedded into age-appropriate cartoons. Visual sensitivity is then quantified through ssVEP amplitudes elicited by a range of different grating frequencies, taking into account looking behaviour. We demonstrate the utility of this approach in a group of subjects with and without IRD (BCVA 0.0–2.0 logMAR) and across a wide age range (1–52 years). We highlight the challenges of this method, as well as its ability to assess individual differences in acuity and quantify treatment-induced improvements. Our protocol requires minimal preparation, bypasses subjective assessments, and is well-suited for both young children and adults facing communicative barriers. We find that this design is not only effective but also engaging, even for our youngest patients. The results will help clinicians and researchers in understanding treatment efficacy and providing critical data for clinical trials and regulatory approval.
Funding & commercial relationships: This work was supported by the NIHR Moorfields Biomedical Research Centre, MeiraGTx/Janssen, and Santen Pharmaceutical Co., Ltd.
Trans-saccadic integration of visual inputs for target recognition peters out with pre-saccadic target eccentricity in visual search
1Eberhard Karls University of Tübingen and Max Planck Institute for Biological Cybernetics, Tübingen, Germany
Bringing a visual object at a peripheral visual location to fovea by a saccade helps recognize this object. Human observers can integrate pre- and post-saccadic visual inputs for the recognition. To our knowledge, this integration has only been studied by an instructed saccade to a target object at a prescribed and known location. Furthermore, the target is typically the only meaningful object in visual display, and the post-saccadic viewing duration of the target is often fixed by experimental design. We study the trans-saccadic integration in visual search behaviour to find and report as soon as possible a target among 404 non-targets. Distributed locations of the search items, each 1.32° × 0.6° in visual angle, spanned 57.3° × 33.8° in visual angle (Liang, Maher, & Zhaoping, 2023 Vision Research 212 108308). Observers decided themselves when and to which locations to make their saccades for the task. We ask whether the pre-saccadic perception may be too ambiguous to contribute to the trans-saccadic integration when the pre-saccadic target is too peripheral from fovea in a crowded scene. To answer this question, we measured (1) the pre-saccadic retinal eccentricity, e, of the target, and (2) the post-saccadic viewing duration, T, which ended when the target was reported. We found that T increased with increasing e and eventually saturated at e around 10°–20°. This is consistent with a trans-saccadic integration that peters out with increasingly peripheral pre-saccadic object, at least when the object perception is highly vulnerable to crowding.
Funding & commercial relationships: This work was supported by funding from the University of Tübingen and the Max Planck Society.
No evidence of cross-orientation suppression differences between migraine aura and healthy controls
NTU Psychology, Nottingham Trent University, UK
Migraine is a common and debilitating neurological disorder, with strong links to visual processing, including photophobia during attacks and differences in visual performance in between attacks. In particular, individuals with migraine with visual aura (MA) experience visual hallucinations prior to the onset of the headache attack. Migraine is thought to be a disorder of inhibitory processing, and this lack of inhibition could be the result of poor communication between the thalamus and the cortex due to disordered oscillatory activity (thalamocortical dysrhythmia; Coppola et al., 2016, Cephalalgia, 36(6), 568–578). Cross-orientation suppression has been used in the past as a measure of inhibition and is thought to originate in the LGN (Li et al., 2006 Journal of Neurophysiology, 96, 1755–1764). We measured cross-orientation suppression and neural oscillations using EEG in MA and control groups. We found no differences in cross-orientation suppression, and no differences in peak resting alpha frequency between migraine and control groups, suggesting that these suppression mechanisms are intact in people who experience migraine aura.
Funding & commercial relationships: CW funded by the Scholarship Projects for Undergraduate Researchers (SPUR) summer internship scheme at NTU
Saccadic eye movements in functional cognitive disorder
SSEHS, Loughborough University, UK
Neurological Disorder characterised by memory or cognition complaints not caused by a systemic or brain disease (McWhirter et al., 2020 Lancet Psychiatry 7 191–207). FCD is distinct from cognitive impairment and does not necessarily indicate the initiation of cognitive decline as few individuals with FCD progress to dementia (Jessen et al., 2020 Lancet Neurol 19 271–278). However, FCD is common in Memory Clinics worldwide (Pennington et al., 2019 J Alzheimers Dis 48 S19–S24). Establishing whether cognitive complaints are functional or systemic in nature is vital for clinicians and objective tests are required. Previous research indicates that participants in the early stages of Alzheimer's disease can be differentiated from healthy controls by measuring antisaccade eye movement deficits (Wilcockson, et al., 2019 Aging 11 5389). Therefore, eye movements may be able to objectively ascertain whether self-reported memory or cognition complaints are functional or systemic in nature. In this study, FCD participants were Memory Clinic patients who self-reported memory complaints but showed no evidence of memory issues on memory tests. FCD participants were compared to a group of patients with Mild Cognitive Impairment and healthy controls on saccadic eye movement tasks. Antisaccade results indicated that FCD differed significantly from MCI but not from healthy controls, but FCD did differ significantly from healthy controls in prosaccades. These results indicate that simple eye movement tasks can differentiate functional from systemic memory issues in patients presenting at Memory Clinics and could aid patient diagnosis.
Funding & commercial relationships: This work was funded by EPSRC project EP/M006255/1 Monitoring Of Dementia using Eye Movements (MODEM)
Visual perceptual memory for crossed and uncrossed binocular disparities
Institute of Physiology, Medical School, University of Pécs, Hungary
Perceptual memory has been investigated for many visual attributes (e.g., spatial frequency, colour, contrast) and these studies revealed specific time-courses and dependence on stimulus parameters (Magnussen & Greenlee, 1999 Psychol Res 62 81–92). We examined visual perceptual memory for binocular depth using Dynamic Random Dot Stereograms (DRDS) with near or far disparities. We employed four different reference disparities in a delayed match-to-sample paradigm at two contrast levels for interstimulus intervals (ISI) up to 4 s. Test stimuli represented a range of equally spaced values centred around the reference disparity of the actual trial. The effect of a masking stimulus during memory retention was also tested. Accuracy of the responses was better for small (+/− 17,5′) than for large (+/−28′) reference disparities and it decayed as a function of ISI. The PSE showed a systematic shift with increasing ISI for both high and low disparities gradually moving towards about 20–23′; and away from the reference disparity. The effect of masking stimuli on the PSE was more pronounced when their disparities differed more from that of the reference. Our results suggest that the retention of absolute disparity is imprecise, and it converges toward a central value, which might represent an optimum in low-level perceptual memory processes, or it could be attributed to perceptual averaging, wherein the visual system computes a statistical summary of the displayed disparities over time. Both mechanisms may point to values that have biological relevance in the temporary storage of depth information for the successful interaction with the constantly changing environment.
Funding & commercial relationships: 2017-1.2.1.-NKP-2017-00002 Hungarian Brain Research Program
Multimodal integration of emotional robotic stimuli: a reaction time study
1School of Psychology, University of Lincoln, UK
2School of Computer Science and Informatics, Cardiff University, UK
Socially Assistive Robots (SARs) are increasingly used as intervention tools for autistic children. However, research into how humans process information from SARs remains sparse. Previous research has shown a facilitation in the processing of human emotional expressions when faces and voices represent the same emotion (congruent) compared to when they represent different emotions (incongruent) or when they are presented only in one modality (unimodal; Collignon et al., 2008 Brain Research 1242 126–135). We investigated whether the processing of auditory, visual and audio-visual SAR stimuli follows similar multisensory processing principles compared to human stimuli. Happy and angry robotic voices and body movements were either presented as unimodal stimuli (voices or bodies) or they were combined to display emotionally congruent (e.g., angry voice-angry body), and incongruent stimuli (e.g., angry voice-happy body). Ninety-five participants were asked to attend either to the body or the voice and to indicate the emotion of the attended modality as fast and as accurately as possible. Inverse efficiency scores and balanced integration scores (Liesefeld & Markus, 2018 Behavior Research Methods 51 40–60) were calculated using the reaction times and accuracy of each condition and participant. Performance was better for emotionally congruent compared to incongruent stimuli, suggesting enhanced processing of congruent stimuli. Moreover, performance was better for unimodal visual than for unimodal auditory stimuli. Results suggest that multisensory emotional SAR stimuli processing reflects that of humans. Further investigation using more varied emotional expressions will give a more detailed understanding of the effects on processing efficiency of different emotions.
Funding & commercial relationships: University of Lincoln
Validity of oculomotor measures of efference copy
School of Psychology, Cardiff University, UK
Predictive processing theories posit that the brain generates predictions of sensory events and compares these predictions to actual experiences. Discrepancies, or prediction errors, can then be used to help guide attention and reinforce learning. An important mechanism in these theories is efference copy, a neural representation of an outgoing motor command that can be used to generate a prediction of the sensory consequences of our own actions (Bansal, S., et al., (2018). The function and failure of sensory predictions. Annals of the New York Academy of Sciences, 1426(1), 199–220.). Many oculomotor tasks purport to quantify the contribution of efference copies to stable visual perception (Thakkar, K. N., Diwadkar, V. A., & Rolfs, M. (2017). Oculomotor prediction: a window into the psychotic mind. Trends in Cognitive Sciences, 21(5), 344–356.), but it is not clear if these tasks are measuring the same underlying mechanism. To investigate this, this study took five oculomotor tasks that have been used to investigate efference copy and ran them in a healthy sample (N=92). There was a smooth pursuit, in-flight displacement, double-step, antisaccade, and perisaccadic mislocalisation task in the battery. If the tasks measure the same underlying mechanism they should all correlate. While there were four significant relationships, many of the correlations reported more evidence in favour of the null hypothesis. This is inconsistent with the tasks measuring a single factor such as efference copy.