Show Reference: "Spatial properties of neurons in the monkey striate cortex"

Spatial properties of neurons in the monkey striate cortex Proceedings of the Royal Society of London. Series B, Containing papers of a Biological character. Royal Society (Great Britain), Vol. 231, No. 1263. (22 July 1987), pp. 251-288 by Michael J. Hawken, Andrew J. Parker
@article{hawken-and-parker-1987,
    abstract = {Contrast sensitivity as a function of spatial frequency was determined for 138 neurons in the foveal region of primate striate cortex. The accuracy of three models in describing these functions was assessed by the method of least squares. Models based on {difference-of-Gaussians} ({DOG}) functions where shown to be superior to those based on the Gabor function or the second differential of a Gaussian. In the most general case of the {DOG} models, each subregion of a simple cell's receptive field was constructed from a single {DOG} function. All the models are compatible with the classical observation that the receptive fields of simple cells are made up of spatially discrete 'on' and 'off' regions. Although the {DOG}-based models have more free parameters, they can account better for the variety of shapes of spatial contrast sensitivity functions observed in cortical cells and, unlike other models, they provide a detailed description of the organization of subregions of the receptive field that is consistent with the physiological constraints imposed by earlier stages in the visual pathway. Despite the fact that the {DOG}-based models have spatially discrete components, the resulting amplitude spectra in the frequency domain describe complex cells just as well as simple cells. The superiority of the {DOG}-based models as a primary spatial filter is discussed in relation to popular models of visual processing that use the Gabor function or the second differential of a Gaussian.},
    author = {Hawken, Michael J. and Parker, Andrew J.},
    day = {22},
    issn = {0080-4649},
    journal = {Proceedings of the Royal Society of London. Series B, Containing papers of a Biological character. Royal Society (Great Britain)},
    keywords = {biology, visual, visual-processing},
    month = jul,
    number = {1263},
    pages = {251--288},
    pmid = {2889214},
    posted-at = {2013-03-19 10:13:17},
    priority = {2},
    title = {Spatial properties of neurons in the monkey striate cortex},
    url = {http://view.ncbi.nlm.nih.gov/pubmed/2889214},
    volume = {231},
    year = {1987}
}

See the CiteULike entry for more info, PDF links, BibTex etc.

Certain receptive fields in the cat striate cortex can be modeled reasonably well using linear filters, more specifically Gabor filters.

The receptive fields of LGN cells can be described as either an excitatory area inside an inhibitory area or the reverse.

The receptive field properties of neurons in the cat striate cortex have been modeled as linear filters. In particular three types of linear filters have been proposed:

  • Gabor filters,
  • filters that based on second differentials of Gaussians functions,
  • difference of Gaussians filters.

Hawken and Parker studied the response patterns of a large number of cells in the cat striate cortex and found that Gabor filters, filters which are second differential of Gaussian functions, and difference-of-Gaussians filters all model these response patterns well, quantitatively.

They found, however, that difference-of-Gaussians filters strongly outperformed the other models.

Difference-of-Gaussians filters are parsimonious candidates for modeling the receptive fields of striate cortex cells, because the kind of differences of Gaussians used in striate cortex (differences of Gaussians with different peak locations) can themselves be computed linearly from differences of Gaussians which model receptive fields of LGN cells (where the peaks coincide), which provide the input to the striate cortex.

X-cells in the cat retina can be modeled by the difference of two Gaussian weighting functions.

Both simple and complex cells' receptive fields can be described using difference-of-Gaussians filters.