Built on the theories of biological neural network, artificial neural network methods have shown many significant advantages. However, the memory space in an artificial neural chip...
This paper outlines a neural network model based on the Fuzzy Cognitive Maps (FCM) framework for solving the automatic image change detection problem. Each pixel in the reference i...
A multilayer neural network based on multivalued neurons (MLMVN) is a neural network with a traditional feedforward architecture. At the same time, this network has a number of spe...
Igor N. Aizenberg, Dmitriy Paliy, Jacek M. Zurada,...
This work presents an implementation of Neocognitron Neural Network, using a high performance computing architecture based on GPU (Graphics Processing Unit). Neocognitron is an ar...
Multi-valued and universal binary neurons (MVN and UBN) are the neural processing elements with the complex-valued weights and high functionality. It is possible to implement an a...