Spatial frequency selectivity of cells in macaque visual cortex.

By R. Valois, D. G. Albrecht, Lisa G. Thorell. 1982

We measured the spatial frequency contrast sensitivity of cells in the primate striate cortex at two different eccentricities to provide quantitative statistics from a large population of cells. Distributions of the peak frequencies and bandwidths are presented and examined in relationship to (a) each other, (b) absolute contrast sensitivity, (c) orientation tuning, (d)retinal eccentricity, and (e) cell type. Simple and complex cells are examined in relationship to linear/nonlinear (that is, X/Y) properties; a procedure is described which provides a simple, reliable and quantitative method for classifying and describing striate cells. Among other things, it is shown that (a) many stirate cells have quite narrow spatial bandwidths and (b) at a given retinal eccentricity, the distribution of peak frequency covers a wide range of frequencies; these findings support the basic multiple channel notion. The orientation tuning and spatial frequency tuning which occurs at the level of striate cortex (in a positively correlated fashion) suggests that the cells might best be considered as two-dimensional spatial filters.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

On the Analysis and Interpretation of Inhomogeneous Quadratic Forms as Receptive Fields (2006, unknown): “In our approach, the quadratic forms are interpreted as second-order approximations of the input-output functions computed by the neurons, and the resulting characterization is similar to the one given by classical physiological experiments (e.g., De Valois, Albrecht, & Thorell, 1982; De Valois, Y”
Fast Feature Pyramids for Object Detection (2014, unknown): “This was suggested by the architecture of the primate visual system [22], where striate cortical cells (roughly equivalent to a wavelet expansion of an image) outnumber retinal ganglion cells (a representation close to image pixels) by a factor ranging from 10 2 to 10 3 .”
Image quality assessment based on a degradation model (2000, unknown): “In the filter bank, we select a one-octave bandwidth to model the the cortical bandpass frequency channels [27].”
Evaluation of deep learning approaches based on convolutional neural networks for corrosion detection (2018, unknown): “Valois et al.9 proposed that human brains operate as two-dimensional (2D) spatial filters in a psychophysical study.”
Realistic Modeling of Simple and Complex Cell Tuning in the HMAX Model, and Implications for Invariant Object Recognition in Cortex (2004, unknown): “For each grating frequency, the maximal cell response was recorded to fit a tuning curve and the spatial frequency selectivity bandwidth was calculated as in [9] by dividing the frequency score at the high crossover of the curve at halfamplitude by the low crossover at the same level.”
Robust Object Recognition with Cortex-Like Mechanisms (2007, methodology): “All filter parameters, i.e., the aspect ratio, ∞ = 0.3, the orientation µ, the effective width ae, the wavelength ‚ as well as the filter sizes s were adjusted so that the tuning properties of the corresponding S1 units match the bulk of V1 parafoveal simple cells based on data from two groups [ 37 ”
Slow feature analysis yields a rich repertoire of complex cell properties. (2005, unknown): “In our case this corresponds to 5150 polynomials of degree 2.”
Development of pattern vision following early and extended blindness (2014, unknown): “In both brain and behavior, contrast sensitivity functions (CSFs) exhibit a characteristic shape: a band-pass function with peak contrast sensitivity and a falloff at relatively lower and higher spatial frequencies.”