The Dynamics of Image Processing by Feature Maps in the Primary Visual Cortex

The operational characteristics of a linear neural network image processing system based on the brain's vision system are investigated. The final stage of the network consists of edge detectors of various orienations arranged in a feature map, corresponding to the primary visual cortex, or V1. The lateral geniculate nucleus is modeled as a preprocessing stage. Excitatory forward and inhibitory backward connections exist between the LGN and V1. By a method of reconstructing the input images in terms of V1 activities, the simulations show that images can be faithfully represented in V1 by the proposed network. The signal-to-noise ratio of the image is improved by the representation, and compression ratios of well over two-hundred are possible. Lateral interacations between V1 neurons sharpen their orientational tuning. We further study the dynamics of the processing, showing that the rate of decrease of the error of the representation is maximized for the receptive fields used, and we develop a Fokker-Planck equation for a more detailed prediction of the error value vs. time. Finally, we show how the eigenvalues of the covariance matrix of the inputs can be employed to predict the rate of error decrease.

[1]  Mats Österberg,et al.  Computing the Karhunen-Loeve Expansion with a Parallel, Unsupervised Filter System , 1992, Neural Computation.

[2]  Kanter,et al.  Eigenvalues of covariance matrices: Application to neural-network learning. , 1991, Physical review letters.

[3]  John G. Daugman,et al.  Complete discrete 2-D Gabor transforms by neural networks for image analysis and compression , 1988, IEEE Trans. Acoust. Speech Signal Process..

[4]  Gyöngyi Gaál,et al.  Population Coding by Neurons in Intrinsic Coordinate Systems 501 tions and to exploit the parallels of this machine vision approach and the tensor network theory of the central nervous system , 2005 .

[5]  Yehoshua Y. Zeevi,et al.  The Generalized Gabor Scheme of Image Representation in Biological and Machine Vision , 1988, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Zhaoping Li,et al.  Toward a Theory of the Striate Cortex , 1994, Neural Computation.

[7]  Ryotaro Kamimura,et al.  Internal representation with minimum entropy in recurrent neural networks: minimizing entropy through inhibitory connections , 1993 .

[8]  Mj Martin Bastiaans.  A Sampling Theorem For The Complex Spectrogram, And Gabor's Expansion Of A Signal In Gaussian Elementary Signals , 1981 .

[9]  H. Barlow,et al.  A model for the economical encoding of the visual image in cerebral cortex , 2004, Biological Cybernetics.

[10]  R. Vogels,et al.  Population coding of stimulus orientation by striate cortical cells , 1990, Biological Cybernetics.

[11]  Robert E. Wyatt,et al.  A computational model of vertical signal propagation in the primary visual cortex , 2004, Biological Cybernetics.

[12]  B J Richmond,et al.  Temporal encoding of two-dimensional patterns by single units in primate primary visual cortex. II. Information transmission. , 1990, Journal of neurophysiology.