An Artificial Neural Network Based on Oxide Synaptic Transistor for Accurate and Robust Image Recognition

Synaptic transistors with low-temperature, solution-processed dielectric films have demonstrated programmable conductance, and therefore potential applications in hardware artificial neural networks for recognizing noisy images. Here, we engineered AlO /InO synaptic transistors via a solution proces...

Full description

Saved in:
Bibliographic Details
Published in:Micromachines (Basel) 2024-04, Vol.15 (4), p.433
Main Authors: Su, Dongyue, Liang, Xiaoci, Geng, Di, Wu, Qian, Liu, Baiquan, Liu, Chuan
Format: Article
Language:English
Subjects:
Aluminum
artificial neural network
Artificial neural networks
Bias
Electrolytes
Hardware
Hydrogen
image recognition
Low temperature
Machine vision
Metal oxides
Neural networks
Silicon wafers
synaptic transistors
Temperature
Thin films
Transistors
Voltage pulses
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Synaptic transistors with low-temperature, solution-processed dielectric films have demonstrated programmable conductance, and therefore potential applications in hardware artificial neural networks for recognizing noisy images. Here, we engineered AlO /InO synaptic transistors via a solution process to instantiate neural networks. The transistors show long-term potentiation under appropriate gate voltage pulses. The artificial neural network, consisting of one input layer and one output layer, was constructed using 9 × 3 synaptic transistors. By programming the calculated weight, the hardware network can recognize 3 × 3 pixel images of characters z, v and n with a high accuracy of 85%, even with 40% noise. This work demonstrates that metal-oxide transistors, which exhibit significant long-term potentiation of conductance, can be used for the accurate recognition of noisy images.
ISSN:2072-666X
2072-666X
DOI:10.3390/mi15040433