▪ Abstract It has long been assumed that sensory neurons are adapted, through both evolutionary and developmental processes, to the statistical properties of the signals to which they are exposed. Attneave (1954), Barlow (1961) proposed that information theory could provide a link between environmental statistics and neural responses through the concept of coding efficiency. Recent developments in statistical modeling, along with powerful computational tools, have enabled researchers to study more sophisticated statistical models for visual images, to validate these models empirically against large sets of data, and to begin experimentally testing the efficient coding hypothesis for both individual neurons and populations of neurons.

/pdf/natural-image-statistics-and-neural-representation-4w6r9kucbm.pdf

Natural image statistics and neural representation

http://vpnl.stanford.edu/papers/kgs_etal_TICS.06.pdf

Repetition and the brain: neural models of stimulus-specific effects

Synaptic plasticity in adult neural circuits may involve the strengthening or weakening of existing synapses as well as structural plasticity, including synapse formation and elimination. Indeed, long-term in vivo imaging studies are beginning to reveal the structural dynamics of neocortical neurons in the normal and injured adult brain. Although the overall cell-specific morphology of axons and dendrites, as well as of a subpopulation of small synaptic structures, are remarkably stable, there is increasing evidence that experience-dependent plasticity of specific circuits in the somatosensory and visual cortex involves cell type-specific structural plasticity: some boutons and dendritic spines appear and disappear, accompanied by synapse formation and elimination, respectively. This Review focuses on recent evidence for such structural forms of synaptic plasticity in the mammalian cortex and outlines open questions.

Experience-dependent structural synaptic plasticity in the mammalian brain.

Bayesian surprise attracts human attention.

To improve the performance of particle image velocimetry in measuring instantaneous velocity fields, direct cross-correlation of image fields can be used in place of auto-correlation methods of interrogation of double- or multiple-exposure recordings. With improved speed of photographic recording and increased resolution of video array detectors, cross-correlation methods of interrogation of successive single-exposure frames can be used to measure the separation of pairs of particle images between successive frames. By knowing the extent of image shifting used in a multiple-exposure and by a priori knowledge of the mean flow-field, the cross-correlation of different sized interrogation spots with known separation can be optimized in terms of spatial resolution, detection rate, accuracy and reliability.

Theory of cross-correlation analysis of PIV images : Image analysis as measuring technique in flows

Complex cells in striate cortex of macaque showed a rapid pattern-specific adaptation. Adaptation made cells more sensitive to orientation change near the adapting orientation. It reduced correlations among the responses of populations of cells, thereby increasing the information transmitted by each action potential. These changes were brought about by brief exposures to stationary patterns, on the time scale of a single fixation. Thus, if successive fixations expose neurons' receptive fields to images with similar but not identical structure, adaptation will remove correlations and improve discriminability.

http://www.cnbc.cmu.edu/~tai/readings/categorization/MulleEtAl1999Science.pdf

Rapid Adaptation in Visual Cortex to the Structure of Images

https://aria.cvs.rochester.edu/papers/roorda_vr2001.pdf

Packing arrangement of the three cone classes in primate retina.

Direct imaging of the retina by adaptive optics allows assessment of the relative number of long-wavelength-sensitive (L) and middle-wavelength-sensitive (M) cones in living human eyes. We examine the functional consequences of variation in the relative numbers of L and M cones (L/M cone ratio) for two observers whose ratios were measured by direct imaging. The L/M cone ratio for the two observers varied considerably, taking on values of 1.15 and 3.79. Two sets of functional data were collected: spectral sensitivity measured with the flicker electroretinogram (ERG) and the wavelength of unique yellow. A genetic analysis was used to determine L and M cone spectra appropriate for each observer. Rayleigh matches confirmed the use of these spectra. We determined the relative strength of L and M cone contributions to ERG spectral sensitivity by fitting the data with a weighted sum of L and M cone spectra. The relative strengths so determined (1.06 and 3.38) were close to the cone ratios established by direct imaging. Thus variation in L/M cone ratio is preserved at the sites tapped by the flicker ERG. The wavelength of unique yellow varied only slightly between the two observers (576.8 and 574.7 nm). This small variation indicates that neural factors play an important role in stabilizing unique yellow against variation in the L/M cone ratio.

/pdf/functional-consequences-of-the-relative-numbers-of-l-and-m-3y5aevgm8i.pdf

Functional consequences of the relative numbers of L and M cones

Functional consequences of the relative numbers of L and M cones (vol 17, pg 607, 2000)

Normal eye movements ensure that the visual world is seen episodically, as a series of often stationary images. In this paper we characterize the responses of neurons in striate cortex to stationary grating patterns presented with abrupt onset. These responses are distinctive. In most neurons the onset of a grating gives rise to a transient discharge that decays with a time constant of 100 msec or less. The early stages of response have higher contrast gain and higher response gain than later stages. Moreover, the variability of discharge during the onset transient is disproportionately low. These factors together make the onset transient an information-rich component of response, such that the detectability and discriminability of stationary gratings grows rapidly to an early peak, within 150 msec of the onset of the response in most neurons. The orientation selectivity of neurons estimated from the first 150 msec of discharge to a stationary grating is indistinguishable from the orientation selectivity estimated from longer segments of discharge to moving gratings. Moving gratings are ultimately more detectable than stationary ones, because responses to the former are continuously renewed. The principal characteristics of the response of a neuron to a stationary grating—the initial high discharge rate, which decays rapidly, and the change of contrast gain with time—are well captured by a model in which each excitatory synaptic event leads to an immediate reduction in synaptic gain, from which recovery is slow.

https://www.jneurosci.org/content/jneuro/21/17/6978.full.pdf

Andrew Metha

Papers

Rapid Adaptation in Visual Cortex to the Structure of Images

Packing arrangement of the three cone classes in primate retina.

Functional consequences of the relative numbers of L and M cones

Functional consequences of the relative numbers of L and M cones (vol 17, pg 607, 2000)

Information Conveyed by Onset Transients in Responses of Striate Cortical Neurons