Eero P. Simoncelli

New York University | NYU · Center for Neural Science (CNS)

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 o...

Internal representations are not uniquely identifiable from perceptual measurements: different representations can generate identical perceptual predictions, and similar representations may predict dissimilar percepts. Here, we generalize a previous method ("Eigendistortions" - Berardino et al., 2017) to enable comparison of models based on their m...

Human ability to discriminate and identify visual attributes varies across the visual field, and is generally worse in the periphery than in the fovea. This decline in performance is revealed in many kinds of tasks, from detection to recognition. A parsimonious hypothesis is that the representation of any visual feature is blurred (spatially averag...

Human ability to discriminate and identify visual attributes varies across the visual field, and is generally worse in the periphery than in the fovea. This decline in performance is revealed in many kinds of tasks, from detection to recognition. A parsimonious hypothesis is that the representation of any visual feature is blurred (spatially averag...

Neurons in early sensory areas rapidly adapt to changing sensory statistics, both by normalizing the variance of their individual responses and by reducing correlations between their responses. Together, these transformations may be viewed as an adaptive form of statistical whitening. Existing mechanistic models of adaptive whitening exclusively us...

The retina transmits visual signals to the brain in the spiking activity of retinal ganglion cells (RGCs). This signal is necessarily imperfect: some visual information is lost in phototransduction and retinal processing. To quantify the transmitted visual signal, we developed a Bayesian method to reconstruct images from the simultaneously recorded...

Sensory systems across all modalities and species exhibit adaptation to continuously changing input statistics. Individual neurons have been shown to modulate their response gains so as to maximize information transmission in different stimulus contexts. Experimental measurements have revealed additional, nuanced sensory adaptation effects includin...

Human ability to discriminate and identify visual attributes varies across the visual field, and is generally worse in the periphery than in the fovea. This decline in performance is revealed in many kinds of tasks, from detection to recognition. A parsimonious hypothesis is that the representation of any visual feature is blurred (spatially averag...

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 liv...

Self-supervised Learning (SSL) provides a strategy for constructing useful representations of images without relying on hand-assigned labels. Many such methods aim to map distinct views of the same scene or object to nearby points in the representation space, while employing some constraint to prevent representational collapse. Here we recast the p...

Observer motion and continuous deformations of objects and surfaces imbue natural videos with distinct temporal structures, enabling partial prediction of future frames from past ones. Conventional methods first estimate local motion, or optic flow, and then use it to predict future frames by warping or copying content. Here, we explore a more dire...

Statistical whitening transformations play a fundamental role in many computational systems, and may also play an important role in biological sensory systems. Individual neurons appear to rapidly and reversibly alter their input-output gains, approximately normalizing the variance of their responses. Populations of neurons appear to regulate their...

Neuroscience has long been an important driver of progress in artificial intelligence (AI). We propose that to accelerate progress in AI, we must invest in fundamental research in NeuroAI.

Perceptual sensitivity often improves with training, a phenomenon known as 'perceptual learning'. Another important perceptual dimension is appearance, the subjective sense of stimulus magnitude. Are training-induced improvements in sensitivity accompanied by more accurate appearance? Here, we examine this question by measuring both discrimination...

A fraction of the visual information arriving at the retina is transmitted to the brain by signals in the optic nerve, and the brain must rely solely on these signals to make inferences about the visual world. Previous work has probed the visual information contained in retinal signals by reconstructing images from retinal activity using linear reg...

The perception of sensory attributes is often quantified through measurements of discriminability (the ability to detect small stimulus changes), as well as through direct judgements of appearance or intensity. Despite their ubiquity, the relationship between these two measurements remains controversial and unresolved. Here, we propose a framework...

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 the...

Denoising is a fundamental challenge in scientific imaging. Deep convolutional neural networks (CNNs) provide the current state of the art in denoising photographic images. However, their potential has been inadequately explored for scientific imaging. Denoising CNNs are typically trained on clean images corrupted with artificial noise, but in scie...

Many sensory-driven behaviors rely on predictions about future states of the environment. Visual input typically evolves along complex temporal trajectories that are difficult to extrapolate. We test the hypothesis that spatial processing mechanisms in the early visual system facilitate prediction by constructing neural representations that follow...

A bstract 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 fMRI, reporting preferred spatial frequencies (tuning curve peaks) of V1 voxels as a function of eccentricity, but their results differ by as much a...

A deep convolutional neural network has been developed to denoise atomic-resolution transmission electron microscope image datasets of nanoparticles acquired using direct electron counting detectors, for applications where the image signal is severely limited by shot noise. The network was applied to a model system of CeO 2 -supported Pt nanopartic...

Sensory processing necessitates discarding some information in service of preserving and reformatting more behaviorally relevant information. Sensory neurons seem to achieve this by responding selectively to particular combinations of features in their inputs, while averaging over or ignoring irrelevant combinations. Here, we expose the perceptual...

Deep convolutional neural networks (CNNs) for image denoising are usually trained on large datasets. These models achieve the current state of the art, but they have difficulties generalizing when applied to data that deviate from the training distribution. Recent work has shown that it is possible to train denoisers on a single noisy image. These...

Significance Humans have a remarkable ability to remember images they have seen, even after seeing thousands, each only once and for a few seconds. One important step toward understanding how the primate brain supports this remarkable form of memory involves pinpointing the neural activity patterns that enable image memory behavior. This paper pres...

Sensory-guided behavior requires reliable encoding of stimulus information in neural responses, and task-specific decoding through selective combination of these responses. The former has been the topic of intensive study, but the latter remains largely a mystery. We propose a framework in which shared stochastic modulation of task- informative neu...

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 comparis...

A deep learning-based convolutional neural network has been developed to denoise atomic-resolution in situ TEM image datasets of catalyst nanoparticles acquired on high speed, direct electron counting detectors, where the signal is severely limited by shot noise. The network was applied to a model catalyst of CeO2-supported Pt nanoparticles. We lev...

Objective measures of image quality generally operate by comparing pixels of a “degraded” image to those of the original. Relative to human observers, these measures are overly sensitive to resampling of texture regions (e.g., replacing one patch of grass with another). Here, we develop the first full-reference image quality model with explicit tol...

Deep convolutional neural networks (CNNs) currently achieve state-of-the-art performance in denoising videos. They are typically trained with supervision, minimizing the error between the network output and ground-truth clean videos. However, in many applications, such as microscopy, noiseless videos are not available. To address these cases, we bu...

Denoising is a fundamental challenge in scientific imaging. Deep convolutional neural networks (CNNs) provide the current state of the art in denoising natural images, where they produce impressive results. However, their potential has barely been explored in the context of scientific imaging. Denoising CNNs are typically trained on real natural im...

Prior probability models are a central component of many image processing problems, but density estimation is notoriously difficult for high-dimensional signals such as photographic images. Deep neural networks have provided state-of-the-art solutions for problems such as denoising, which implicitly rely on a prior probability model of natural imag...

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 ignor...

We develop a model for representing visual texture in a low-dimensional feature space, along with a novel self-supervised learning objective that is used to train it on an unlabeled database of texture images. Inspired by the architecture of primate visual cortex, the model uses a first stage of oriented linear filters (corresponding to cortical ar...

Neural populations do not perfectly encode the sensory world: their capacity is limited by the number of neurons, metabolic and other biophysical resources, and intrinsic noise. The brain is presumably shaped by these limitations, improving efficiency by discarding some aspects of incoming sensory streams, while preferentially preserving commonly o...

The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human judgments. Perceptual datasets (e.g., LIVE and TID2013) gathered for this purpose provide useful benchmarks for improving IQA methods, but their heavy use creates a risk of overfitting. Here, we perform a large-sca...

Objective measures of image quality generally operate by making local comparisons of pixels of a "degraded" image to those of the original. Relative to human observers, these measures are overly sensitive to resampling of texture regions (e.g., replacing one patch of grass with another). Here we develop the first full-reference image quality model...

Responses of sensory neurons are often modeled using a weighted combination of rectified linear subunits. Since these subunits often cannot be measured directly, a flexible method is needed to infer their properties from the responses of downstream neurons. We present a method for maximum likelihood estimation of subunits by soft-clustering spike-t...

Motion selectivity in primary visual cortex (V1) is approximately separable in orientation, spatial frequency, and temporal frequency ("frequency-separable"). Models for area MT neurons posit that their selectivity arises by combining direction-selective V1 afferents whose tuning is organized around a tilted plane in the frequency domain, specifyin...

Motion selectivity in primary visual cortex (V1) is approximately separable in orientation, spatial frequency, and temporal frequency (“frequency-separable”). Models for area MT neurons posit that their selectivity arises by combining direction-selective V1 afferents whose tuning is organized around a tilted plane in the frequency domain, specifyin...

Deep convolutional networks often append additive constant ("bias") terms to their convolution operations, enabling a richer repertoire of functional mappings. Biases are also used to facilitate training, by subtracting mean response over batches of training images (a component of "batch normalization"). Recent state-of-the-art blind denoising meth...

Many behaviors rely on predictions derived from recent visual input, but the temporal evolution of those inputs is generally complex and difficult to extrapolate. We propose that the visual system transforms these inputs to follow straighter temporal trajectories. To test this ‘temporal straightening’ hypothesis, we develop a methodology for estima...

The original and corrected figures are shown in the accompanying Author Correction.

Sensory-guided behavior requires reliable encoding of information (from stimuli to neural responses) and flexible decoding (from neural responses to behavior). In typical decision tasks, a small subset of cells within a large population encode task-relevant stimulus information and need to be identified by later processing stages for relevant infor...

Integration of rectified synaptic inputs is a widespread nonlinear motif in sensory neuroscience. We present a novel method for maximum likelihood estimation of nonlinear subunits by soft-clustering spike-triggered stimuli. Subunits estimated from parasol ganglion cells recorded in macaque retina partitioned the receptive field into compact regions...

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...

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 represe...

We develop a method for comparing hierarchical image representations in terms of their ability to explain perceptual sensitivity in humans. Specifically, we utilize Fisher information to establish a model-derived prediction of sensitivity to local perturbations around a given natural image. For a given image, we compute the eigenvectors of the Fish...

We compare several functional models of LGN population response in terms of their ability to predict human judgments of visual distortion. The model-derived Fisher Information matrix provides a bound on discrimination thresholds for the visibility of arbitrary distortions. In particular, the largest and smallest eigenvectors of this matrix represen...

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-selec...

We develop a framework for rendering photographic images, taking into account display limitations, so as to optimize perceptual similarity between the rendered image and the original scene. We formulate this as a constrained optimization problem, in which we minimize a measure of perceptual dissimilarity, the Normalized Laplacian Pyramid Distance (...

We describe an image compression method, consisting of a nonlinear analysis transformation, a uniform quantizer, and a nonlinear synthesis transformation. The transforms are constructed in three successive stages of convolutional linear filters and nonlinear activation functions. Unlike most convolutional neural networks, the joint nonlinearity is...

We describe an image compression system, consisting of a nonlinear encoding transformation, a uniform quantizer, and a nonlinear decoding transformation. Like many deep neural network architectures, the transforms consist of layers of convolutional linear filters and nonlinear activation functions, but we use a joint nonlinearity that implements a...

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. H...

We introduce a general framework for end-to-end optimization of the rate–distortion performance of nonlinear transform codes assuming scalar quantization. The framework can be used to optimize any differentiable pair of analysis and synthesis transforms in combination with any differentiable perceptual metric. As an example, we consider a code buil...

Significance The brain generates increasingly complex representations of the visual world to recognize objects, to form new memories, and to organize visual behavior. Relatively simple signals in the retina are transformed through a cascade of neural computations into highly complex responses in visual cortical areas deep in the temporal lobe. The...