Optical inner-product implementation of neural networks models.

The Hopfield neural network model is described in terms of inner product, i.e., as a matched filtering step followed by a pattern synthesis step. Optical implementation by two cascaded coherent filtering setups with holographic matched filters is described. Suitable encoding of information in the form of bipolar (positive or negative) amplitudes in one hologram and of non-negative amplitudes in the other allows one to deal only with non-negative quantities in the input and output planes, thereby avoiding use of multiple channels and coherent detection. The performance of this scheme is analytically evaluated. The above coding can moreover be adapted to the cases of nonzero average memorized states and to the higher-order models associated to Hopfield's algorithm. Numerical simulations and experimental results are presented to illustrate the analysis.

[1]  J J Hopfield,et al.  Neurons with graded response have collective computational properties like those of two-state neurons. , 1984, Proceedings of the National Academy of Sciences of the United States of America.

[2]  A W Lohmann,et al.  On Moiré fringes as Fourier test objects. , 1966, Applied optics.

[3]  A F Gmitro,et al.  Hopfield model associative memory with nonzero-diagonal terms in memory matrix. , 1988, Applied optics.

[4]  S Lowenthal,et al.  Optical Fourier transform: what is the optimal setup? , 1982, Applied optics.

[5]  H H Arsenault,et al.  Modification of the threshold condition for a contentaddressable memory based on the Hopfield model. , 1987, Applied optics.

[6]  Demetri Psaltis,et al.  Nonlinear discriminant functions and associative memories , 1987 .

[7]  D Psaltis,et al.  Optical information processing based on an associative-memory model of neural nets with thresholding and feedback. , 1985, Optics letters.

[8]  A. B. Vander Lugt,et al.  Signal detection by complex spatial filtering , 1964, IEEE Trans. Inf. Theory.

[9]  Y Owechko,et al.  Optoelectronic resonator neural networks. , 1987, Applied optics.

[10]  Demetri Psaltis,et al.  Optical Associative Memory Using Fourier Transform Holograms , 1987 .

[11]  Pierre Chavel,et al.  A proposed generalization of Hopfield's algorithm , 1987 .

[12]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[13]  G. J. Dunning,et al.  Holographic associative memory with nonlinearities in the correlation domain. , 1987, Applied optics.

[14]  R A Athale,et al.  Optical implementation of associative memory with controlled nonlinearity in the correlation domain. , 1986, Optics letters.

[15]  Demetri Psaltis,et al.  Shift-Invariant Optical Associative Memories , 1987 .

[16]  D Psaltis,et al.  Optical implementation of the Hopfield model. , 1985, Applied optics.

[17]  A. Lohmann,et al.  Complex spatial filtering with binary masks. , 1966, Applied optics.