Conference Papers Year : 2022

Spiking Neuron Network Based on VTEAM Memristor and MOSFET-LIF Neuron

Abstract

Neuromorphic computing has been widely developed due to its low power consumption and powerful interpretability. LIF neurons, the general-purpose neurons in neuromorphic computing, are under constant research in hardware implementations of spiking neural networks. In this paper, we design a LIF circuit with MOSFET based on the mathematical model of the LIF neuron. The simulated circuit can be directly applied to the spiking neural network through the VTEAM memristor crossbar architecture. The effect of parameter changes in the circuit on the membrane potential is demonstrated. Finally, we validate feasibility of the process on the DVS128 gesture dataset using a generic spiking neural network architecture and obtain satisfactory performance.
Fichier principal
Vignette du fichier
537972_1_En_5_Chapter.pdf (1.91 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04666435 , version 1 (01-08-2024)

Licence

Identifiers

Cite

Jiahui Fu, Shuiping Gou, Zhang Guo. Spiking Neuron Network Based on VTEAM Memristor and MOSFET-LIF Neuron. 5th International Conference on Intelligence Science (ICIS), Oct 2022, Xi'an, China. pp.37-44, ⟨10.1007/978-3-031-14903-0_5⟩. ⟨hal-04666435⟩
20 View
1 Download

Altmetric

Share

More