A neuromorphic impulsive circuit for processing dynamic signals

The electronic implementation of a highly parallel neuromorphic circuit which processes and classifies dynamic signals is described. The authors discuss the silicon implementation of an artificial dendritic tree that has a spatiotemporal processing capability that is modeled after biological neurons which have extensive passive dendritic branching. Background material is presented on biological neuron behavior that is important in processing dynamic signals. The architecture and circuit details of artificial dendrites and artificial synapses are presented. The properties of both active and passive biological dendritic trees as well as the dynamic and static behavior of chemical synapses were studied. A simple and scalable electronic circuit implemented in standard CMOS technology is described.<<ETX>>

[1]  J. G. Elias,et al.  Silicon implementation of an artificial dendritic tree , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[2]  M. Berry,et al.  The application of network analysis to the study of branching patterns of large dendritic fields , 1976, Brain Research.

[3]  Carver Mead,et al.  Analog VLSI and neural systems , 1989 .

[4]  J. G. Elias,et al.  A generic algorithm for training networks with artificial dendritic trees , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[5]  C. Mead,et al.  The silicon retina. , 1991 .

[6]  J. G. Elias.  Spatial-temporal properties of artificial dendritic trees , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.