This paper investigates the architectural requirements in simulating large neuralnetworks using a highly parallel multiprocessor with distributed memory and optical interconnects. First, we model the structure of a n...
详细信息
The neural net model introduced by Hogg and Huberman1,2 is examined and extended to include 2-dimensional inputs. Computer simulations of this model exhibit fault tolerant behavior in pattern recognition applications ...
详细信息
Many articles on neuralnetworks focus on learning, and restrict themselves to a limited class of simple neurons. The present paper is a tutorial designed, by contrast to emphasize the "domain-specific" stru...
详细信息
The authors describe their investigation of electronic implementations of fine-grained parallel computingmodels that are loosely drawn from models of biological neural function. Experimental custom chips that combine...
详细信息
The authors describe their investigation of electronic implementations of fine-grained parallel computingmodels that are loosely drawn from models of biological neural function. Experimental custom chips that combine a new mix of analog and digital processing with standard fabrication technology have shown the feasibility of the neuralnetwork approach. Early results on transforming aspects of biological computing to electronic hardware suggest that networks of highly-interconnected, simple, low-precision processors may provide new tools for tackling problems that have been difficult for standard computers.
A learning algorithm based on temporal difference of membrane potential of the neuron is proposed for self-organizing neuralnetworks. It is independent of the neuron nonlinearity, so it can be applied to analog or bi...
详细信息
Various N 2 weighted reconfigurable opticalnetworks for the optical free-space either neuralnetwork (ONN) or other chip-to-chip interconnections are proposed. With these new schemes, the large number of either elect...
详细信息
Various N 2 weighted reconfigurable opticalnetworks for the optical free-space either neuralnetwork (ONN) or other chip-to-chip interconnections are proposed. With these new schemes, the large number of either electronic or optical delay lines, that were used in the previous ONN implementations, are minimized, so that the massive interconnections among monolithically intergrated either optical neurons or computing chips can be implemented. Both linearly- and circularly-distributed optical neuron (chip) interconnect models are presented. A comparison between the two interconnect models is also given.
The neuralcomputing scheme of image reconstruction by the human visual system has been modeled by multi-scale zero-crossings as unique representations of bandlimited polynomial functions. The exact analytical develop...
详细信息
A hardware implementation of a lightly connected artificial neuralnetwork known as the Hogg-Huberman model (1) (2) is described. The hardware is built around NCR's Geometric Arithmetic Parallel Processor (GAPP) c...
详细信息
Variation in the response of hardware components causes analog implementations of neuralnetworks with soft nonlinearities to corrupt their signals with noise. In this paper we simulate the behavior of analog implemen...
详细信息
optical processor architectures for various forms of the alternating projection neuralnetwork (APNN) are considered. Required iteration is performed by passive optical feedback using only free space and guided propag...
详细信息
暂无评论