During last years a lot of attention have been focused to the hardware implementation of Artificial Neural Networks (ANN) to efficiently exploit the inherent parallelism associated to these systems. From the different types of ANN, the Spiking Neural Networks (SNN) arise as a promising bio-inspired model that is able to emulate the expected neural behavior with a high confidence. Many works are centered in using analog circuitry to reproduce SNN with a high degree of precision, while minimizing the area and the energy costs. Nevertheless, the reliability and flexibility of these systems is lower if compared with digital implementations. In this paper we present a new, low-cost bio-inspired digital neural model for SNN along with an auxiliary Computer Aided Design (CAD) tool for the efficient implementation of high-volume SNN.
CITATION STYLE
Galán-Prado, F., & Rosselló, J. L. (2017). Smart hardware implementation of spiking neural networks. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 10305 LNCS, 560–568. https://doi.org/10.1007/978-3-319-59153-7_48
Mendeley helps you to discover research relevant for your work.