A CORF computational model of a simple cell that relies on LGN input outperforms the Gabor function model

Abstract

Simple cells in primary visual cortex are believed to extract local contour information from a visual scene. The 2D Gabor function (GF) model has gained particular popularity as a computational model of a simple cell. However, it short-cuts the LGN, it cannot reproduce a number of properties of real simple cells, and its effectiveness in contour detection tasks has never been compared with the effectiveness of alternative models. We propose a computational model that uses as afferent inputs the responses of model LGN cells with center-surround receptive fields (RFs) and we refer to it as a Combination of Receptive Fields (CORF) model. We use shifted gratings as test stimuli and simulated reverse correlation to explore the nature of the proposed model. We study its behavior regarding the effect of contrast on its response and orientation bandwidth as well as the effect of an orthogonal mask on the response to an optimally oriented stimulus. We also evaluate and compare the performances of the CORF and GF models regarding contour detection, using two public data sets of images of natural scenes with associated contour ground truths. The RF map of the proposed CORF model, determined with simulated reverse correlation, can be divided in elongated excitatory and inhibitory regions typical of simple cells. The modulated response to shifted gratings that this model shows is also characteristic of a simple cell. Furthermore, the CORF model exhibits cross orientation suppression, contrast invariant orientation tuning and response saturation. These properties are observed in real simple cells, but are not possessed by the GF model. The proposed CORF model outperforms the GF model in contour detection with high statistical confidence (RuG data set: p < 10 −4 , and

Contour Detection by CORF Operator

“…In [2] we show that the proposed CORF operator can also be considered as a computational model of a simple cell. There we show that the multiplicative character of the computation we use to combine model LGN responses is essential to achieve three important properties, namely cross-orientation suppression, contrast invariant orientation tuning and response saturation, which are typical of simple cells.…”

An Electrical Insulator Defects Detection Method Combined Human Receptive Field Model

“…Both of these cells are arranged and their total number is substantially equal and detect contrast changes. Some researchers have proved that the on-type and off-type channels are located between the LGN and the visual cortex is fully parallel separation [21,22]. The process of LGN receptive field is defined as a 2D Gaussian function in the following:…”

Automatic Differentiation of u- and n-serrated Patterns in Direct Immunofluorescence Images

“…A B -COSFIRE filter 3 [3] is a ridge detector, which is based on the COSFIRE approach [2] and the CORF computational model [1]. Its response is achieved by computing the geometric mean of a group of linearly aligned responses of a Difference-of-Gaussians (DoG) filter.…”

“…Such a filter has three parameters: standard deviation σ of the outer Gaussian function in the involved DoG filter 4 , radius l and orientation θ. The radius l is the farthest distance from the center of the filter 1 We use the following parameters: standard deviation of √ 2, high threshold of 0.02 and low threshold of 0.01. 2 We choose the upper-most boundary by comparing the mean of the y-coordinates of the two boundaries 3 Matlab scripts: http://www.mathworks.com/matlabcentral/fileexchange/49172 4 The standard deviation of the inner Gaussian function is 0.5σ at which DoG responses are considered as input to a B -COSFIRE filter in a specific position, Fig.…”

“…We comment on the choice of the values of σ and l in Sections 3 and 4. For further technical details we refer to [1][2][3][7][8][9] and to an online implementation 6 . Fig.…”