An assembly neural network for texture segmentation

-An architecture o f a neural network with assembly organization is described. Such network architecture is applied to the problem of texture segmentation in natural scenes. The network is partitioned into several subnetworks. Each subnetwork is a column structure in which features are represented by means o f "float" coding. Input data excite corresponding "floats'" o f neurons in the subnetworks. In the process o f learning the weights o f modifiable connections between excited neurons are changed so that Hebb "s assemblies are formed in the column structures. All subnetworks are incorporated into a single network by a neural activity control system. Computer simulation o f the proposed network has been performed. The results o f computer simulations show the possibility o f successful application o f the assembly neural network to the problem o f texture segmentation. Copyright ©1996 Elsevier Science Ltd Keywords--Neural assembly, Neural columns, Floats of neurons, Disjoint subnetworks, Texture segmentation.