Lightness constancy-the ability to form a stable percept of surface reflectance despite changes in illumination and context-is critical for the accurate interpretation of visual scenes, including virtual content in optical see-through augmented reality (OST-AR). However, the additive blending of virtual and real-world scenes in AR elevates the luminance of stimuli, which can make it difficult to create a convincing visual impression of low-reflectance surfaces. Here we investigated how a range of visual cues affects lightness perception in OST-AR, with the goal of identifying viewing conditions that expand the range of perceived reflectance for virtual stimuli. In two experiments, participants viewed virtual test patches in a foreground plane and rated their lightness relative to real patches on a paper reference palette. We manipulated four factors: the composition of surrounding elements in the foreground plane, the content and luminance of the background plane, the motion of the background plane, and the relative disparity between foreground and background. We found that the strongest effects on perceived reflectance came from changes to the foreground context: Isolating the test patch or eliminating luminance variation in the surrounding texture substantially reduced the range of perceived reflectance. Surprisingly, even large changes to the background, such as removing texture, introducing motion, or eliminating the depth disparity between foreground and background layers, had much smaller effects. These findings show that lightness perception in OST-AR depends heavily on how the foreground is structured and that including a rich foreground context can help virtual surfaces to generate a wide and realistic range of perceived reflectance.

Fast and efficient reading involves not only processing the currently fixated word but also simultaneous processing of the upcoming text. Although the role of extrafoveal processing in natural reading has been extensively documented, its contribution to reading acceleration benefits has remained unexplored. In the present study, we sought to investigate whether reading acceleration might act by influencing the efficiency and spatial extent of parafoveal processing. Participants read at either a natural or an accelerated pace while the availability of parafoveal information was independently manipulated for the first (N+1) and second (N+2) words beyond fixation. Results showed a significant increase in reading speed during accelerated reading, accompanied by more strategic oculomotor behaviors. Importantly, fixation durations analyzed across valid and invalid (random-letters masking) preview conditions revealed an increased sensitivity to orthographic irregularities in parafoveal vision at greater eccentricities from fixation during accelerated reading. Moreover, participants who exhibited greater parafoveal preview effects demonstrated a more pronounced increase in reading speed. These results demonstrate that, during accelerated reading, readers flexibly adjust the spatial allocation of their limited processing resources, strategically widening the window from which extrafoveal information is extracted. Furthermore, they offer new insights into reading efficiency and promising avenues for targeted interventions in poor or impaired readers.

Mathematical modeling of visual textures traces back to Julesz's intuition that texture perception in humans is based on local correlations between image features. An influential approach for texture analysis and generation generalizes this notion to linear correlations between the nonlinear features computed by convolutional neural networks (CNNs), compiled into Gram matrices. Given that CNNs are often used as models for the visual system, it is natural to ask whether such "texture representations" spontaneously align with the textures' perceptual content, and, in particular, whether those CNNs that are regarded as better models for the visual system also possess more human-like texture representations. Here we quantify the perceptual content captured by feature correlations computed for a diverse pool of CNNs, and we compare it to the models' perceptual alignment with the mammalian visual system as measured by Brain-Score. Surprisingly, we find that there is no connection between conventional measures of CNN quality as a model of the visual system and its alignment with human texture perception. We conclude that texture perception involves mechanisms that are distinct from those that are commonly modeled using approaches based on CNNs trained on object recognition, possibly depending on the integration of contextual information.

Contrast determines how strong visual stimuli drive neural responses and, in turn, how well we can see. This contrast-response relationship is nonlinear, with gain control mechanisms acting to compress responses to high-contrast stimuli, effectively setting the contrast range that neural responses are most sensitive to. Yet, capturing these saturating responses noninvasively in human functional magnetic resonance imaging (fMRI) has proven challenging. Here, we introduce a model-based fMRI approach to more flexibly and efficiently estimate saturating contrast responses: population contrast response function (CRF) mapping. To vet this technique, we benchmarked pCRF mapping against traditional deconvolved BOLD responses for stimuli presented with a fast event-related design. We found tight correspondence in parameter estimates between the deconvolution-derived CRF and pCRF estimates, validating the model-based approach. We then asked whether pCRF mapping could push beyond conventional paradigms by employing a continuous, highly condition-rich stimulation design that affords greater efficiency and flexibility but does not lend itself to traditional fMRI analyses. As predicted by the model-based approach, voxel-wise pCRF estimates could still be recovered from this unconventional design. In sum, we demonstrated that model-based pCRF mapping dissociates CRF estimation from strict design constraints and broadens the space of usable experimental paradigms while still capturing the saturating nonlinearities that are fundamental to neural population responses.

Under mesopic conditions, such as dawn or dusk, signals from both rod and cone photoreceptors contribute to perception. These parallel inputs are combined within the retina before being sent to subsequent visual areas. The integration of these kinetically distinct parallel signals poses unique challenges for human vision. Though previous behavioral studies have found that dim lighting conditions specifically impair motion perception in human subjects, the origin of this dependence is unclear. In the present study, we created a model circuit that predicts ganglion cell responses to moving stimuli by incorporating electrophysiologically derived circuit components into a Hassenstein-Reichardt correlator, a classical motion-detection model. The model circuit demonstrates that interactions between rod- and cone-derived signals negatively impact the encoding of the direction of a moving object under mesopic conditions. Furthermore, we found that motion discriminability improved if the model circuit was only sensitive to the cone-activating components of the stimuli. We conclude that rod+cone signal interference occurring at the lowest level of vision has an impact on motion direction discrimination, a higher-level task with relevance for behavior.

Effective data visualization and statistics education both depend on helping viewers detect meaningful patterns in graphs. Here we investigated pattern detection in 2 × 2 graphs, a widely used form of data visualization. The 2 × 2 graphs display a continuous dependent variable across two predictors, each with two levels. Factor A appears on the x-axis. Factor B appears as distinct visual series, such as black versus white lines or bars. Prior research on 2 × 2 graphs examined viewers' verbal descriptions of Factor A, Factor B, and A × B interaction patterns. Those studies revealed what people can describe but not what they can perceive. The present experiment addressed that gap by testing how viewers learn to perceptually organize Factor A, Factor B, and interaction patterns in 2 × 2 graphs. In a preregistered trial-and-error task, 416 adults learned to classify 2 × 2 line or bar graphs according to Factor A, Factor B, or A × B interaction patterns. The line and bar graphs contained identical numeric information. Two primary findings emerged. First, line graphs produced higher accuracy than bar graphs for interaction and Factor B patterns. No such advantage appeared for Factor A patterns, where accuracy remained low across graph types. Second, participants who trained on bar graphs often adopted a flawed rule. They reported Factor A and Factor B patterns only when an interaction appeared. This classification error did not occur for line graphs. Both primary findings matched preregistered predictions. These results show that data visualization choices, including graph type and predictor placement, influence the perceptual availability of patterns in 2 × 2 graphs. Clarifying those perceptual limits advances both data visualization design and statistics education.

Visual search theories distinguish between attentional capture and disengagement, yet how multiple feature dimensions combine to delay disengagement remains unclear. This study investigated whether shared features (color, shape, object category) produce additive or super-additive costs in oculomotor disengagement. Forty-three participants performed a delayed-disengagement task where a central fixated stimulus shared zero to three features with a peripheral target. Saccadic latency (SL) served as the index of disengagement cost. Results showed that SL increased nonlinearly with the number of shared features, following a quadratic trend. A component decomposition revealed that this acceleration was driven primarily by color: Color overlap produced large main effects, whereas shape and object-category overlap produced costs primarily when combined with color. Our findings suggest that disengagement is not a simple summation of feature similarities. Instead, color may act as a high-priority gate: A color match may open a capacity-limited verification stage in which each additional shared feature disproportionately prolongs inspection. We interpret these findings in terms of a candidate hierarchical gating account, in which high-salience features such as color may govern entry into verification.

The gulf between mathematically motivated stimuli-such as sine waves and noise ensembles-and images drawn from natural scenes may not be as wide as it appears. For the former, the mathematical properties that support their utility as analytic probes are explicit. For naturalistic stimuli, this is not the case, but they are created by physical processes whose outcomes-at least in principle-heavily constrain their statistics. Images generated computationally occupy a middle ground: Their mathematical properties remain implicit in the rules that generate them, but these rules are simpler than the processes that generate natural images. It is suggested that a greater understanding of the statistics of natural images, while challenging, will enhance their utility as experimental tools and thereby lead to new insights into the design principles of sensory systems.

Human pupillary responses are influenced by not only low-level visual properties but also cognitive and affective factors related to task demands. Nevertheless, how the temporal dynamics of pupil modulation differ across evaluative tasks in the context of material perception remains unclear. In this study, we investigated how pupillary responses vary when observers evaluate the same set of object images for either glossiness or attractiveness. These two perceptual attributes were selected to represent distinct cognitive demands: Glossiness is more closely related to lower-level visual processing, whereas attractiveness involves higher-level evaluative processing. The stimuli included 60 grayscale object photographs selected from the THINGS database, representing common real-world items without social or facial content. Low-level image statistics were controlled using histogram matching to equalize luminance and contrast across all the stimuli. Participants viewed each image for 3,000 ms while maintaining central fixation and rated its glossiness or attractiveness on a 7-point scale in separate task blocks. Grand-average waveforms revealed task-dependent modulations of pupil size across rating levels, which is consistent with previous reports that pupillary responses vary with evaluative context even for identical stimuli. Furthermore, temporal principal component analysis and generalized additive modeling revealed that higher glossiness ratings were associated with greater pupil constriction at early, light reflex-like latencies, whereas higher attractiveness ratings elicited greater pupil dilation at later time points. These findings suggest that distinct temporal profiles of pupil size reflect task-specific processing demands in material perception, which is potentially consistent with differences in temporal weighting between visual and affective evaluations.

When searching for objects, our sensory resources are limited. According to the zoom lens model of attention, the visual system can flexibly adapt the size of the attentional spotlight applied to an array of items depending on task difficulty. When search tasks are relatively straightforward, attention can be distributed widely across space to process several items simultaneously. However, when discrimination becomes more difficult, the attentional spotlight constricts to concentrate resources to a limited area resulting in the processing of only one or very few items simultaneously at any one time. The present study aimed to empirically test the zoom lens account by assessing search performance while systematically altering the visible area in the search display (field of view [FOV]). We presented two search tasks of contrasting difficulty in a virtual reality environment with a head-contingent display. When the FOV became smaller, there were no differences in search times between the two search tasks, a result which seems to contradict the predictions of the zoom lens model. The zoom lens model predicts that restricted viewing in a demanding search task should exact a smaller cost because of the presence of the smaller attentional window. However, we did not find any evidence of such an effect. Results from an analysis of head and eye movements also suggested that when using both fixations and head movements to scan their environment, participants used head movements to reveal new areas of the visual scene in a manner largely disconnected from classic eye movement and attentional processes. Once the head came to rest, attentional shifts and eye movements were used to search within the newly revealed area.

A complete empirical characterization of color discrimination in three dimensions has long remained out of reach. Classical studies, beginning with MacAdam's ellipses, provided local measurements in restricted chromatic planes, but a spatially dense and internally consistent mapping of discrimination structure across full color space has not yet been achieved. Here, we present such a systematic three-dimensional measurement of color discrimination in the standardized sRGB stimulus space. Eight observers measured discrimination regions at 35 reference colors distributed on a body-centered cubic lattice within the RGB cube. At each location, color differences were probed along seven orientations, yielding 14 directional extents. These measurements defined centrally symmetric convex regions that were fitted with minimum-volume ellipsoids, providing a compact description of local discrimination structure. Ellipsoids were represented as symmetric positive-definite matrices and analyzed using a Frobenius geometry, enabling normalization across observers and smooth interpolation to arbitrary locations. The resulting metric field is spatially smooth, highly structured, and remarkably consistent across observers up to an individual global scale factor. Grain size increases along the achromatic axis and exhibits systematic chromatic asymmetries. Comparison with CIEDE2000 reveals substantial agreement in overall scale variation but systematic differences in local anisotropy. Because the metric field is defined geometrically, it can be transformed into linear RGB or any other color representation through the corresponding coordinate transformations. Together, these data provide a coherent three-dimensional empirical mapping of color discrimination across a large ecologically relevant volume of color space and establish an empirical framework for perceptual color metrics.

Free-viewing perimetry is a continuous psychophysical paradigm that aims to assess visual field defects (VFD) while relying solely on an analysis of recorded gaze. Such an approach could benefit patients who cannot be assessed using standard automated perimetry. Previous research has demonstrated that gaze behavior during the free viewing of movie clips contains information on the integrity of the visual field. Yet, thus far, no single analytical approach could reconstruct all types of (simulated) VFD. In the current study, we take a distinctive analytical approach towards VFD reconstruction by explicitly taking into account events that were not gazed at, rather than primarily focussing on those that were. To demonstrate the potential of this approach, we re-analyzed previously recorded data from a gaze-contingent simulated VFD experiment. We compared a participant's gaze behavior to that of control participants while they watched different movie clips. In our analysis, we give specific weight to the events that were prioritized by the control participants, yet missed out upon by those with simulated defects. Our findings indicate that this approach allows us to reliably detect the presence of, and to varying degrees reconstruct five types of simulated visual field defects. We conclude that free-viewing perimetry is feasible and may form the foundation for an inclusive complement to standard automated perimetry.

Accurately mapping receptive fields under naturalistic viewing requires correcting for distortions introduced by eye movements. Current approaches often rely on parameterized "shifter networks" optimized with Poisson generalized linear models, but these methods are computationally intensive and limited to small regions of interest. We present a new framework that exploits the translation properties of Hartley basis functions to model eye position effects directly as phase shifts, eliminating the need for explicit retinal image reconstruction. This formulation admits closed-form gradients, enabling efficient parameter optimization and rapid convergence. Validation on simulated data shows that the method accurately recovers both ground-truth receptive fields and the underlying image transformation. Applied to high-density mouse V1 recordings, the approach substantially improves receptive field sharpness compared to naive estimates. While the method is specific to Hartley basis stimuli, once calibrated, it provides a reusable mapping between eye position and retinal translation. This efficiency and scalability make the technique a practical tool for receptive field mapping in free-viewing experiments and for integration with optical imaging and large-scale electrophysiology. We discuss the relative advantages, limitations, and applicability of the method in comparison with existing convolutional neural network/shifter-network approaches.

Multisensory integration is often assumed to increase when visual input is degraded, yet it remains unclear whether low vision enhances susceptibility to cross-modal illusions or whether such effects depend on local variations in visual reliability. We tested low-vision and sighted control participants on the sound-induced flash illusion (SIFI) across 24 visual-field locations, concurrently measuring baseline flash-detection accuracy. While both groups showed the expected auditory-driven increase in perceived flash numerosity, sighted controls were significantly more susceptible to the classic "double-flash" percept. While illusion strength approached marginal significance for eccentricity, it was more robustly predicted by local flash-detection accuracy, indicating that sound-induced percepts depend on the availability of a reliable visual signal. Exploratory Bayesian causal inference modeling revealed that when fitted across all locations, model performance was systematically weaker and more variable for low-vision observers. Crucially, however, restricting the analysis to high-visibility locations yielded robust model fits by the Bayesian information criterion and close agreement between observed and predicted response proportions. This suggests that, like sighted controls, low-vision participants continue to behave in a manner broadly consistent with Bayesian-optimal inference when visual information is sufficiently reliable. Together, these findings demonstrate that low vision does not globally amplify audiovisual interactions; rather, multisensory enhancements are constrained by local visual reliability, and Bayesian-like inference remains most evident when provided with a sufficient threshold of sensory input.

This study investigated attentional templates and color tuning during efficient visual search. To evaluate optimal tuning and relational accounts, we used a paradigm where participants searched for a color singleton target while ignoring a cross-dimensional shape distractor. Crucially, the distractor's color varied along a precise continuum within the same dimension, spanning values toward and away from the nontarget background. We modeled response times (RTs) and oculomotor capture by fitting asymmetric Gaussian functions to the response curves. Supporting optimal tuning, oculomotor capture (Experiment 1) and manual responses under fixed fixation (Experiment 2) revealed peak interference systematically shifted away from the nontarget color to maximize target-nontarget contrast. Intriguingly, this peak shift was significantly more pronounced in early oculomotor capture than manual RTs, suggesting that optimal tuning biases peripheral attentional guidance (guiding template) more strongly than foveated target verification (target template). Furthermore, response functions exhibited profound asymmetry, displaying shallower slopes for colors extending along the target-defining vector. Whereas Experiment 3 confirmed highly efficient search slopes, Experiment 4 demonstrated that a feature-dissimilar distractor produced negligible capture, validating that selection was governed by a feature-based template. Together, these findings indicate that although attentional templates center on a specifically tuned, contrast-maximizing peak, their overall distribution remains bound to a relative, directional feature rule, consistent with the relational account. Data and scripts are available at https://doi.org/q9f4.