跳到主要导航 跳到搜索 跳到主要内容

Artificial Optoelectronic Synapses Based on TiNxO2–x/MoS2 Heterojunction for Neuromorphic Computing and Visual System

  • Wenxiao Wang
  • , Song Gao
  • , Yang Li*
  • , Wenjing Yue
  • , Hao Kan
  • , Chunwei Zhang
  • , Zheng Lou
  • , Lili Wang
  • , Guozhen Shen*
  • *此作品的通讯作者
  • University of Jinan
  • CAS - Institute of Semiconductors

科研成果: 期刊稿件 › 文章 › 同行评审

摘要

Being capable of dealing with both electrical signals and light, artificial optoelectronic synapses are of great importance for neuromorphic computing and are receiving a burgeoning amount of interest in visual information processing. In this work, an artificial optoelectronic synapse composed of Al/TiNxO2–x/MoS2/ITO (H-OSD) is proposed and experimentally realized. The H-OSD can enable basic electrical voltage-induced synaptic functions such as the long/short-term plasticity and moreover the synaptic plasticity can be electrically adjusted. In response to the light stimuli, versatile advanced synaptic functions including long/short-term memory, and learning-forgetting-relearning are successfully demonstrated, which could enhance the information processing capability for neuromorphic computing. Most importantly, based on these light-induced salient features, a 4 × 4 synapse array is developed to show the potential application of the proposed H-OSD in constructing artificial visual system. It is shown that the perceiving and memorizing of the light information that are respectively relevant to the visual perception and visual memory functions, can be readily attained through tuning of the light intensity and the number of illuminations. As such, the proposed optoelectronic synapse shows great potentials in both neuromorphic computing and visual information processing and will facilitate the applications such as electronic eyes and light-driven neurorobotics.

源语言英语
期刊论文编号2101201
期刊Advanced Functional Materials
卷31
期34
DOI
出版状态已出版 - 20 8月 2021
已对外发布是

学术指纹

探究 'Artificial Optoelectronic Synapses Based on TiNxO2–x/MoS2 Heterojunction for Neuromorphic Computing and Visual System' 的科研主题。它们共同构成独一无二的学术指纹。

引用此