A two-layer neural peak detector

Investigation of a single-layer peak detector based on the Hopfield-Tank architecture revealed that the network cannot be implemented in hardware using existing VLSI technology once the network exceeds approximately 150 neurons. It is shown that using a two-layer scheme reduces the number of neuron interconnections and thus makes hardware implementation of large networks feasible.<<ETX>>

[1]  G. L. Dempsey,et al.  Hough transform system enhancements resulting from neural network implementation , 1991, IEEE Proceedings of the SOUTHEASTCON '91.

[2]  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.

[3]  R.M. Inigo,et al.  Artificial neural network implementation of the hough transform , 1989, Twenty-Third Asilomar Conference on Signals, Systems and Computers, 1989..

[4]  Jerome A. Feldman,et al.  Connectionist Models and Their Properties , 1982, Cogn. Sci..

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