There has been a strong push recently to examine biological scale simulationsof neuromorphic algorithms to achieve stronger inference capabilities. Thispaper presents a set of piecewise linear spiking neuron models, which canreproduce different behaviors, similar to the biological neuron, both for asingle neuron as well as a network of neurons. The proposed models areinvestigated, in terms of digital implementation feasibility and costs,targeting large scale hardware implementation. Hardware synthesis and physicalimplementations on FPGA show that the proposed models can produce preciseneural behaviors with higher performance and considerably lower implementationcosts compared with the original model. Accordingly, a compact structure of themodels which can be trained with supervised and unsupervised learningalgorithms has been developed. Using this structure and based on a spike ratecoding, a character recognition case study has been implemented and tested.
展开▼
