pRAM nets for detection of small targets in sequences of infra-red images
Neural Networks - Special issue: automatic target recognition
Learning Probabilistic RAM Nets Using VLSI Structures
IEEE Transactions on Computers
IWANN '96 Proceedings of the International Workshop on Artificial Neural Networks: From Natural to Artificial Neural Computation
Hi-index | 0.00 |
The pRAM has been implemented as a VLSI processor incorporating 256 neurons with on-chip learning. Several such processors may be connected to form larger networks. The pRAM was originally conceived to model the noisy release of neurotransmitter vesicles in synaptic connections. In a network, it forms n-tuples and may operate as a noisy lookup table with generalization. The reinforcement training used is also biologically realistic and has a scope ranging from a single neuron to the whole network, the architecture being user defined.