Search

Visual processing necessitates both extracting and discarding information. Here, the authors use a specialized set of stimuli and two complementary discrimination tasks to demonstrate the opposing perceptual implications of these two aspects of information processing.

The authors examined neuronal responses in V1 and V2 to synthetic texture stimuli that replicate higher-order statistical dependencies found in natural images. V2, but not V1, responded differentially to these textures, in both macaque (single neurons) and human (fMRI). Human detection of naturalistic structure in the same images was predicted by V2 responses, suggesting a role for V2 in representing natural image structure.

The authors developed a model of neuron firing in which spike generation arises from the combination of sensory drive and stimulus-independent modulatory influences. This model provides an accurate account of neuron responses in multiple visual areas, suggesting that variability originates from excitability fluctuations that increase in strength along the visual pathway.

Speed preferences in MT neurons are found to be unaffected by changes in stimulus pattern, supporting the hypothesis that these neurons represent retinal image velocities.

Animals respond rapidly and precisely to a variety of sensory stimuli, but the neural mechanisms supporting this flexibility are not fully understood. Here the authors describe a model of adaptive sensory processing based on functionally-targeted stochastic modulation, and find evidence for this co-variability in macaque V1 and middle temporal area.

The visual signals transmitted by the retina to the brain are affected by random drift in eye position, but the impact of this on visual capabilities is not clear. Here, the authors show that the decoding of images from evoked spike trains recorded in the macaque retina improves with fixational eye movements, even when the eye position is unknown.

The brain predicts future sensory input. The authors hypothesize that the visual system achieves this by straightening the temporal trajectories of natural videos, and they provide evidence using human perceptual experiments and computational modeling.

Many behaviours depend on predictions about the environment. Here the authors find neural populations in primary visual cortex to straighten the temporal trajectories of natural video clips, facilitating the extrapolation of past observations.

Orientation judgments are more accurate at the horizontal and vertical orientations, possibly reflecting a statistical inference. Here the authors provide evidence for this idea, finding that observers' internal models for orientation match the local orientation distribution measured in photographs, and suggest how such information could be encoded in a neural population.

Sensory signals are transduced at high resolution, but their structure must be stored in a more compact format. Here the authors show that the auditory system summarizes the temporal details of sounds using time-averaged statistics. Such statistical representations produce good categorical discrimination, but limit the ability to discern temporal detail.

Receptive fields of visual neurons get bigger along the ventral visual pathway and, in each area, they grow with distance from the fovea. The authors exploit these properties to build a model for visual representation in the ventral stream, using 'metameric' visual stimuli (which appear perceptually identical, but are actually different) to test the model predictions. The model can also explain deficits in peripheral recognition known as visual crowding.

The functional significance of correlated firing in a complete population of macaque parasol retinal ganglion cells using a model of multi-neuron spike responses is analysed. Fitting the physiological data to a model of multi-neuron spike responses, it is found that a significant fraction of what is usually considered single-cell noise in trial-to-trial response variability can be explained by correlations, and that a significant amount of sensory information can be decoded from the correlation structure.

One of the ambitions of computational neuroscience is that we will continue to make improvements in the field of artificial intelligence that will be informed by advances in our understanding of how the brains of various species evolved to process information. To that end, here the authors propose an expanded version of the Turing test that involves embodied sensorimotor interactions with the world as a new framework for accelerating progress in artificial intelligence.