A unified framework for perceived magnitude and discriminability of sensory stimuli.

The perception of sensory attributes is often quantified through measurements of sensitivity (the ability to detect small stimulus changes), as well as through direct judgments of appearance or intensity. Despite their ubiquity, the relationship between these two measurements remains controversial and unresolved. Here, we propose a framework in which they arise from different aspects of a common representation. Specifically, we assume that judgments of stimulus intensity (e.g., as measured through rating scales) reflect the mean value of an internal representation, and sensitivity reflects a combination of mean value and noise properties, as quantified by the statistical measure of Fisher information. Unique identification of these internal representation properties can be achieved by combining measurements of sensitivity and judgments of intensity. As a central example, we show that Weber's law of perceptual sensitivity can coexist with Stevens' power-law scaling of intensity ratings (for all exponents), when the noise amplitude increases in proportion to the representational mean. We then extend this result beyond the Weber's law range by incorporating a more general and physiology-inspired form of noise and show that the combination of noise properties and sensitivity measurements accurately predicts intensity ratings across a variety of sensory modalities and attributes. Our framework unifies two primary perceptual measurements-thresholds for sensitivity and rating scales for intensity-and provides a neural interpretation for the underlying representation.

Pinpointing the neural signatures of single-exposure visual recognition memory.

Memories of the images that we have seen are thought to be reflected in the reduction of neural responses in high-level visual areas such as inferotemporal (IT) cortex, a phenomenon known as repetition suppression (RS). We challenged this hypothesis with a task that required rhesus monkeys to report whether images were novel or repeated while ignoring variations in contrast, a stimulus attribute that is also known to modulate the overall IT response. The monkeys' behavior was largely contrast invariant, contrary to the predictions of an RS-inspired decoder, which could not distinguish responses to images that are repeated from those that are of lower contrast. However, the monkeys' behavioral patterns were well predicted by a linearly decodable variant in which the total spike count was corrected for contrast modulation. These results suggest that the IT neural activity pattern that best aligns with single-exposure visual recognition memory behavior is not RS but rather sensory referenced suppression: reductions in IT population response magnitude, corrected for sensory modulation.

Mapping spatial frequency preferences across human primary visual cortex.

Neurons in primate visual cortex (area V1) are tuned for spatial frequency, in a manner that depends on their position in the visual field. Several studies have examined this dependency using functional magnetic resonance imaging (fMRI), reporting preferred spatial frequencies (tuning curve peaks) of V1 voxels as a function of eccentricity, but their results differ by as much as two octaves, presumably owing to differences in stimuli, measurements, and analysis methodology. Here, we characterize spatial frequency tuning at a millimeter resolution within the human primary visual cortex, across stimulus orientation and visual field locations. We measured fMRI responses to a novel set of stimuli, constructed as sinusoidal gratings in log-polar coordinates, which include circular, radial, and spiral geometries. For each individual stimulus, the local spatial frequency varies inversely with eccentricity, and for any given location in the visual field, the full set of stimuli span a broad range of spatial frequencies and orientations. Over the measured range of eccentricities, the preferred spatial frequency is well-fit by a function that varies as the inverse of the eccentricity plus a small constant. We also find small but systematic effects of local stimulus orientation, defined in both absolute coordinates and relative to visual field location. Specifically, peak spatial frequency is higher for pinwheel than annular stimuli and for horizontal than vertical stimuli.

Comparison of Full-Reference Image Quality Models for Optimization of Image Processing Systems.

The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human quality judgments. Perceptual datasets gathered for this purpose have provided useful benchmarks for improving IQA methods, but their heavy use creates a risk of overfitting. Here, we perform a large-scale comparison of IQA models in terms of their use as objectives for the optimization of image processing algorithms. Specifically, we use eleven full-reference IQA models to train deep neural networks for four low-level vision tasks: denoising, deblurring, super-resolution, and compression. Subjective testing on the optimized images allows us to rank the competing models in terms of their perceptual performance, elucidate their relative advantages and disadvantages in these tasks, and propose a set of desirable properties for incorporation into future IQA models.

Selectivity and tolerance for visual texture in macaque V2.

As information propagates along the ventral visual hierarchy, neuronal responses become both more specific for particular image features and more tolerant of image transformations that preserve those features. Here, we present evidence that neurons in area V2 are selective for local statistics that occur in natural visual textures, and tolerant of manipulations that preserve these statistics. Texture stimuli were generated by sampling from a statistical model, with parameters chosen to match the parameters of a set of visually distinct natural texture images. Stimuli generated with the same statistics are perceptually similar to each other despite differences, arising from the sampling process, in the precise spatial location of features. We assessed the accuracy with which these textures could be classified based on the responses of V1 and V2 neurons recorded individually in anesthetized macaque monkeys. We also assessed the accuracy with which particular samples could be identified, relative to other statistically matched samples. For populations of up to 100 cells, V1 neurons supported better performance in the sample identification task, whereas V2 neurons exhibited better performance in texture classification. Relative to V1, the responses of V2 show greater selectivity and tolerance for the representation of texture statistics.

Dissociation of Choice Formation and Choice-Correlated Activity in Macaque Visual Cortex.

Responses of individual task-relevant sensory neurons can predict monkeys' trial-by-trial choices in perceptual decision-making tasks. Choice-correlated activity has been interpreted as evidence that the responses of these neurons are causally linked to perceptual judgments. To further test this hypothesis, we studied responses of orientation-selective neurons in V1 and V2 while two macaque monkeys performed a fine orientation discrimination task. Although both animals exhibited a high level of neuronal and behavioral sensitivity, only one exhibited choice-correlated activity. Surprisingly, this correlation was negative: when a neuron fired more vigorously, the animal was less likely to choose the orientation preferred by that neuron. Moreover, choice-correlated activity emerged late in the trial, earlier in V2 than in V1, and was correlated with anticipatory signals. Together, these results suggest that choice-correlated activity in task-relevant sensory neurons can reflect postdecision modulatory signals.SIGNIFICANCE STATEMENT When observers perform a difficult sensory discrimination, repeated presentations of the same stimulus can elicit different perceptual judgments. This behavioral variability often correlates with variability in the activity of sensory neurons driven by the stimulus. Traditionally, this correlation has been interpreted as suggesting a causal link between the activity of sensory neurons and perceptual judgments. More recently, it has been argued that the correlation instead may originate in recurrent input from other brain areas involved in the interpretation of sensory signals. Here, we call both hypotheses into question. We show that choice-related activity in sensory neurons can be highly variable across observers and can reflect modulatory processes that are dissociated from perceptual decision-making.

Contextual modulation of sensitivity to naturalistic image structure in macaque V2.

The stimulus selectivity of neurons in V1 is well known, as is the finding that their responses can be affected by visual input to areas outside of the classical receptive field. Less well understood are the ways selectivity is modified as signals propagate to visual areas beyond V1, such as V2. We recently proposed a role for V2 neurons in representing the higher order statistical dependencies found in images of naturally occurring visual texture. V2 neurons, but not V1 neurons, respond more vigorously to "naturalistic" images that contain these dependencies than to "noise" images that lack them. In this work, we examine the dependency of these effects on stimulus size. For most V2 neurons, the preference for naturalistic over noise stimuli was modest when presented in small patches and gradually strengthened with increasing size, suggesting that the mechanisms responsible for this enhanced sensitivity operate over regions of the visual field that are larger than the classical receptive field. Indeed, we found that surround suppression was stronger for noise than for naturalistic stimuli and that the preference for large naturalistic stimuli developed over a delayed time course consistent with lateral or feedback connections. These findings are compatible with a spatially broad facilitatory mechanism that is absent in V1 and suggest that a distinct role for the receptive field surround emerges in V2 along with sensitivity for more complex image structure. NEW & NOTEWORTHY The responses of neurons in visual cortex are often affected by visual input delivered to regions of the visual field outside of the conventionally defined receptive field, but the significance of such contextual modulations are not well understood outside of area V1. We studied the importance of regions beyond the receptive field in establishing a novel form of selectivity for the statistical dependencies contained in natural visual textures that first emerges in area V2.

Slow gain fluctuations limit benefits of temporal integration in visual cortex.

Sensory neurons represent stimulus information with sequences of action potentials that differ across repeated measurements. This variability limits the information that can be extracted from momentary observations of a neuron's response. It is often assumed that integrating responses over time mitigates this limitation. However, temporal response correlations can reduce the benefits of temporal integration. We examined responses of individual orientation-selective neurons in the primary visual cortex of two macaque monkeys performing an orientation-discrimination task. The signal-to-noise ratio of temporally integrated responses increased for durations up to a few hundred milliseconds but saturated for longer durations. This was true even when cells exhibited little or no adaptation in their response levels. These observations are well explained by a statistical response model in which spikes arise from a Poisson process whose stimulus-dependent rate is modulated by slow, stimulus-independent fluctuations in gain. The response variability arising from the Poisson process is reduced by temporal integration, but the slow modulatory nature of variability due to gain fluctuations is not. Slow gain fluctuations therefore impose a fundamental limit on the benefits of temporal integration.

Perceptually optimized image rendering.

We develop a framework for rendering photographic images by directly optimizing their perceptual similarity to the original visual scene. Specifically, over the set of all images that can be rendered on a given display, we minimize the normalized Laplacian pyramid distance (NLPD), a measure of perceptual dissimilarity that is derived from a simple model of the early stages of the human visual system. When rendering images acquired with a higher dynamic range than that of the display, we find that the optimization boosts the contrast of low-contrast features without introducing significant artifacts, yielding results of comparable visual quality to current state-of-the-art methods, but without manual intervention or parameter adjustment. We also demonstrate the effectiveness of the framework for a variety of other display constraints, including limitations on minimum luminance (black point), mean luminance (as a proxy for energy consumption), and quantized luminance levels (halftoning). We show that the method may generally be used to enhance details and contrast, and, in particular, can be used on images degraded by optical scattering (e.g., fog). Finally, we demonstrate the necessity of each of the NLPD components-an initial power function, a multiscale transform, and local contrast gain control-in achieving these results and we show that NLPD is competitive with the current state-of-the-art image quality metrics.

A Convolutional Subunit Model for Neuronal Responses in Macaque V1.

The response properties of neurons in the early stages of the visual system can be described using the rectified responses of a set of self-similar, spatially shifted linear filters. In macaque primary visual cortex (V1), simple cell responses can be captured with a single filter, whereas complex cells combine a set of filters, creating position invariance. These filters cannot be estimated using standard methods, such as spike-triggered averaging. Subspace methods like spike-triggered covariance can recover multiple filters but require substantial amounts of data, and recover an orthogonal basis for the subspace in which the filters reside, rather than the filters themselves. Here, we assume a linear-nonlinear-linear-nonlinear (LN-LN) cascade model in which the first LN stage consists of shifted ("convolutional") copies of a single filter, followed by a common instantaneous nonlinearity. We refer to these initial LN elements as the "subunits" of the receptive field, and we allow two independent sets of subunits, each with its own filter and nonlinearity. The second linear stage computes a weighted sum of the subunit responses and passes the result through a final instantaneous nonlinearity. We develop a procedure to directly fit this model to electrophysiological data. When fit to data from macaque V1, the subunit model significantly outperforms three alternatives in terms of cross-validated accuracy and efficiency, and provides a robust, biologically plausible account of receptive field structure for all cell types encountered in V1. SIGNIFICANCE STATEMENT: We present a new subunit model for neurons in primary visual cortex that significantly outperforms three alternative models in terms of cross-validated accuracy and efficiency, and provides a robust and biologically plausible account of the receptive field structure in these neurons across the full spectrum of response properties.

Neural Quadratic Discriminant Analysis: Nonlinear Decoding with V1-Like Computation.

Linear-nonlinear (LN) models and their extensions have proven successful in describing transformations from stimuli to spiking responses of neurons in early stages of sensory hierarchies. Neural responses at later stages are highly nonlinear and have generally been better characterized in terms of their decoding performance on prespecified tasks. Here we develop a biologically plausible decoding model for classification tasks, that we refer to as neural quadratic discriminant analysis (nQDA). Specifically, we reformulate an optimal quadratic classifier as an LN-LN computation, analogous to "subunit" encoding models that have been used to describe responses in retina and primary visual cortex. We propose a physiological mechanism by which the parameters of the nQDA classifier could be optimized, using a supervised variant of a Hebbian learning rule. As an example of its applicability, we show that nQDA provides a better account than many comparable alternatives for the transformation between neural representations in two high-level brain areas recorded as monkeys performed a visual delayed-match-to-sample task.

Near-optimal integration of orientation information across saccades.

We perceive a stable environment despite the fact that visual information is essentially acquired in a sequence of snapshots separated by saccadic eye movements. The resolution of these snapshots varies-high in the fovea and lower in the periphery-and thus the formation of a stable percept presumably relies on the fusion of information acquired at different resolutions. To test if, and to what extent, foveal and peripheral information are integrated, we examined human orientation-discrimination performance across saccadic eye movements. We found that humans perform best when an oriented target is visible both before (peripherally) and after a saccade (foveally), suggesting that humans integrate the two views. Integration relied on eye movements, as we found no evidence of integration when the target was artificially moved during stationary viewing. Perturbation analysis revealed that humans combine the two views using a weighted sum, with weights assigned based on the relative precision of foveal and peripheral representations, as predicted by ideal observer models. However, our subjects displayed a systematic overweighting of the fovea, relative to the ideal observer, indicating that human integration across saccades is slightly suboptimal.

Efficient sensory encoding and Bayesian inference with heterogeneous neural populations.

The efficient coding hypothesis posits that sensory systems maximize information transmitted to the brain about the environment. We develop a precise and testable form of this hypothesis in the context of encoding a sensory variable with a population of noisy neurons, each characterized by a tuning curve. We parameterize the population with two continuous functions that control the density and amplitude of the tuning curves, assuming that the tuning widths vary inversely with the cell density. This parameterization allows us to solve, in closed form, for the information-maximizing allocation of tuning curves as a function of the prior probability distribution of sensory variables. For the optimal population, the cell density is proportional to the prior, such that more cells with narrower tuning are allocated to encode higher-probability stimuli and that each cell transmits an equal portion of the stimulus probability mass. We also compute the stimulus discrimination capabilities of a perceptual system that relies on this neural representation and find that the best achievable discrimination thresholds are inversely proportional to the sensory prior. We examine how the prior information that is implicitly encoded in the tuning curves of the optimal population may be used for perceptual inference and derive a novel decoder, the Bayesian population vector, that closely approximates a Bayesian least-squares estimator that has explicit access to the prior. Finally, we generalize these results to sigmoidal tuning curves, correlated neural variability, and a broader class of objective functions. These results provide a principled embedding of sensory prior information in neural populations and yield predictions that are readily testable with environmental, physiological, and perceptual data.

Spatio-temporal correlations and visual signalling in a complete neuronal population.

Statistical dependencies in the responses of sensory neurons govern both the amount of stimulus information conveyed and the means by which downstream neurons can extract it. Although a variety of measurements indicate the existence of such dependencies, their origin and importance for neural coding are poorly understood. Here we analyse the functional significance of correlated firing in a complete population of macaque parasol retinal ganglion cells using a model of multi-neuron spike responses. The model, with parameters fit directly to physiological data, simultaneously captures both the stimulus dependence and detailed spatio-temporal correlations in population responses, and provides two insights into the structure of the neural code. First, neural encoding at the population level is less noisy than one would expect from the variability of individual neurons: spike times are more precise, and can be predicted more accurately when the spiking of neighbouring neurons is taken into account. Second, correlations provide additional sensory information: optimal, model-based decoding that exploits the response correlation structure extracts 20% more information about the visual scene than decoding under the assumption of independence, and preserves 40% more visual information than optimal linear decoding. This model-based approach reveals the role of correlated activity in the retinal coding of visual stimuli, and provides a general framework for understanding the importance of correlated activity in populations of neurons.

Foundations of visual form selectivity for neurons in macaque V1 and V2.

We have measured the visually evoked activity of single neurons recorded in areas V1 and V2 of awake, fixating macaque monkeys, and captured their responses with a common computational model. We used a stimulus set composed of "droplets" of localized contrast, band-limited in orientation and spatial frequency; each brief stimulus contained a random superposition of droplets presented in and near the mapped receptive field. We accounted for neuronal responses with a 2-layer linear-nonlinear model, representing each receptive field by a combination of orientation- and scale-selective filters. We fit the data by jointly optimizing the model parameters to enforce sparsity and to prevent overfitting. We visualized and interpreted the fits in terms of an "afferent field" of nonlinearly combined inputs, dispersed in the 4 dimensions of space and spatial frequency. The resulting fits generally give a good account of the responses of neurons in both V1 and V2, capturing an average of 40% of the explainable variance in neuronal firing. Moreover, the resulting models predict neuronal responses to image families outside the test set, such as gratings of different orientations and spatial frequencies. Our results offer a common framework for understanding processing in the early visual cortex, and also demonstrate the ways in which the distributions of neuronal responses in V1 and V2 are similar but not identical.

Neuronal and behavioral responses to naturalistic texture images in macaque monkeys.

The visual world is richly adorned with texture, which can serve to delineate important elements of natural scenes. In anesthetized macaque monkeys, selectivity for the statistical features of natural texture is weak in V1, but substantial in V2, suggesting that neuronal activity in V2 might directly support texture perception. To test this, we investigated the relation between single cell activity in macaque V1 and V2 and simultaneously measured behavioral judgments of texture. We generated stimuli along a continuum between naturalistic texture and phase-randomized noise and trained two macaque monkeys to judge whether a sample texture more closely resembled one or the other extreme. Analysis of responses revealed that individual V1 and V2 neurons carried much less information about texture naturalness than behavioral reports. However, the sensitivity of V2 neurons, especially those preferring naturalistic textures, was significantly closer to that of behavior compared with V1. The firing of both V1 and V2 neurons predicted perceptual choices in response to repeated presentations of the same ambiguous stimulus in one monkey, despite low individual neural sensitivity. However, neither population predicted choice in the second monkey. We conclude that neural responses supporting texture perception likely continue to develop downstream of V2. Further, combined with neural data recorded while the same two monkeys performed an orientation discrimination task, our results demonstrate that choice-correlated neural activity in early sensory cortex is unstable across observers and tasks, untethered from neuronal sensitivity, and thus unlikely to reflect a critical aspect of the formation of perceptual decisions. Significance statement: As visual signals propagate along the cortical hierarchy, they encode increasingly complex aspects of the sensory environment and likely have a more direct relationship with perceptual experience. We replicate and extend previous results from anesthetized monkeys differentiating the selectivity of neurons along the first step in cortical vision from area V1 to V2. However, our results further complicate efforts to establish neural signatures that reveal the relationship between perception and the neuronal activity of sensory populations. We find that choice-correlated activity in V1 and V2 is unstable across different observers and tasks, and also untethered from neuronal sensitivity and other features of nonsensory response modulation.

Efficient coding of spatial information in the primate retina.

Sensory neurons have been hypothesized to efficiently encode signals from the natural environment subject to resource constraints. The predictions of this efficient coding hypothesis regarding the spatial filtering properties of the visual system have been found consistent with human perception, but they have not been compared directly with neural responses. Here, we analyze the information that retinal ganglion cells transmit to the brain about the spatial information in natural images subject to three resource constraints: the number of retinal ganglion cells, their total response variances, and their total synaptic strengths. We derive a model that optimizes the transmitted information and compare it directly with measurements of complete functional connectivity between cone photoreceptors and the four major types of ganglion cells in the primate retina, obtained at single-cell resolution. We find that the ganglion cell population exhibited 80% efficiency in transmitting spatial information relative to the model. Both the retina and the model exhibited high redundancy (~30%) among ganglion cells of the same cell type. A novel and unique prediction of efficient coding, the relationships between projection patterns of individual cones to all ganglion cells, was consistent with the observed projection patterns in the retina. These results indicate a high level of efficiency with near-optimal redundancy in visual signaling by the retina.

Targeted V1 comodulation supports task-adaptive sensory decisions.

Sensory-guided behavior requires reliable encoding of stimulus information in neural populations, and flexible, task-specific readout. The former has been studied extensively, but the latter remains poorly understood. We introduce a theory for adaptive sensory processing based on functionally-targeted stochastic modulation. We show that responses of neurons in area V1 of monkeys performing a visual discrimination task exhibit low-dimensional, rapidly fluctuating gain modulation, which is stronger in task-informative neurons and can be used to decode from neural activity after few training trials, consistent with observed behavior. In a simulated hierarchical neural network model, such labels are learned quickly and can be used to adapt downstream readout, even after several intervening processing stages. Consistently, we find the modulatory signal estimated in V1 is also present in the activity of simultaneously recorded MT units, and is again strongest in task-informative neurons. These results support the idea that co-modulation facilitates task-adaptive hierarchical information routing.

Fixational Eye Movements Enhance the Precision of Visual Information Transmitted by the Primate Retina.

Fixational eye movements alter the number and timing of spikes transmitted from the retina to the brain, but whether these changes enhance or degrade the visual signal is unclear. To quantify this, we developed a Bayesian method for reconstructing natural images from the recorded spikes of hundreds of macaque retinal ganglion cells (RGCs) of the major cell types, combining a likelihood model for RGC light responses with the natural image prior implicitly embedded in an artificial neural network optimized for denoising. The method matched or surpassed the performance of previous reconstruction algorithms, and provided an interpretable framework for characterizing the retinal signal. Reconstructions were improved with artificial stimulus jitter that emulated fixational eye movements, even when the jitter trajectory was inferred from retinal spikes. Reconstructions were degraded by small artificial perturbations of spike times, revealing more precise temporal encoding than suggested by previous studies. Finally, reconstructions were substantially degraded when derived from a model that ignored cell-to-cell interactions, indicating the importance of stimulus-evoked correlations. Thus, fixational eye movements enhance the precision of the retinal representation.

Catalyzing next-generation Artificial Intelligence through NeuroAI.

Neuroscience has long been an essential driver of progress in artificial intelligence (AI). We propose that to accelerate progress in AI, we must invest in fundamental research in NeuroAI. A core component of this is the embodied Turing test, which challenges AI animal models to interact with the sensorimotor world at skill levels akin to their living counterparts. The embodied Turing test shifts the focus from those capabilities like game playing and language that are especially well-developed or uniquely human to those capabilities - inherited from over 500 million years of evolution - that are shared with all animals. Building models that can pass the embodied Turing test will provide a roadmap for the next generation of AI.