Implementation of Convolutional Kernel Function Using 3-D TiOx Resistive Switching Devices for Image Processing

Myonghoon Kwak; Jaesung Park; Jiyong Woo; Hyunsang Hwang

doi:10.1109/TED.2018.2862139

Implementation of Convolutional Kernel Function Using 3-D TiO_x Resistive Switching Devices for Image Processing

Myonghoon Kwak
, Jaesung Park
, Jiyong Woo
, Hyunsang Hwang

School of Electronics Engineering

Pohang University of Science and Technology

Research output: Contribution to journal › Article › peer-review

20 Scopus citations

Abstract

We implemented a convolution operation in a convolutional neural network by introducing a 3-D integrated TiO_x-based resistive switching device (RSD) array to act as the kernel. The 3-D integrated TiO_x-based RSD exhibits a gradual SET/RESET behavior, which enables multilevel characteristics. A pair of layers in the 3-D structured array is used to achieve the positive and negative weights. Based on these results, we demonstrated the 3 × 3 Laplace kernel for edge detection using images from the Columbia Object Image Library data set in a simulator to validate the convolution operation.

Original language	English
Article number	8435959
Pages (from-to)	4716-4718
Number of pages	3
Journal	IEEE Transactions on Electron Devices
Volume	65
Issue number	10
DOIs	https://doi.org/10.1109/TED.2018.2862139
State	Published - Oct 2018

Keywords

3-D integrated array
kernel
multilevel
resistive switching device (RSD)
TiO

Access to Document

10.1109/TED.2018.2862139

Cite this

@article{55ca012b9268455087d2736f8989babd,

title = "Implementation of Convolutional Kernel Function Using 3-D TiOx Resistive Switching Devices for Image Processing",

abstract = "We implemented a convolution operation in a convolutional neural network by introducing a 3-D integrated TiOx-based resistive switching device (RSD) array to act as the kernel. The 3-D integrated TiOx-based RSD exhibits a gradual SET/RESET behavior, which enables multilevel characteristics. A pair of layers in the 3-D structured array is used to achieve the positive and negative weights. Based on these results, we demonstrated the 3 × 3 Laplace kernel for edge detection using images from the Columbia Object Image Library data set in a simulator to validate the convolution operation.",

keywords = "3-D integrated array, kernel, multilevel, resistive switching device (RSD), TiO",

author = "Myonghoon Kwak and Jaesung Park and Jiyong Woo and Hyunsang Hwang",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.",

year = "2018",

month = oct,

doi = "10.1109/TED.2018.2862139",

language = "English",

volume = "65",

pages = "4716--4718",

journal = "IEEE Transactions on Electron Devices",

issn = "0018-9383",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "10",

}

TY - JOUR

T1 - Implementation of Convolutional Kernel Function Using 3-D TiOx Resistive Switching Devices for Image Processing

AU - Kwak, Myonghoon

AU - Park, Jaesung

AU - Woo, Jiyong

AU - Hwang, Hyunsang

PY - 2018/10

Y1 - 2018/10

N2 - We implemented a convolution operation in a convolutional neural network by introducing a 3-D integrated TiOx-based resistive switching device (RSD) array to act as the kernel. The 3-D integrated TiOx-based RSD exhibits a gradual SET/RESET behavior, which enables multilevel characteristics. A pair of layers in the 3-D structured array is used to achieve the positive and negative weights. Based on these results, we demonstrated the 3 × 3 Laplace kernel for edge detection using images from the Columbia Object Image Library data set in a simulator to validate the convolution operation.

AB - We implemented a convolution operation in a convolutional neural network by introducing a 3-D integrated TiOx-based resistive switching device (RSD) array to act as the kernel. The 3-D integrated TiOx-based RSD exhibits a gradual SET/RESET behavior, which enables multilevel characteristics. A pair of layers in the 3-D structured array is used to achieve the positive and negative weights. Based on these results, we demonstrated the 3 × 3 Laplace kernel for edge detection using images from the Columbia Object Image Library data set in a simulator to validate the convolution operation.

KW - 3-D integrated array

KW - kernel

KW - multilevel

KW - resistive switching device (RSD)

KW - TiO

UR - https://www.scopus.com/pages/publications/85051660706

U2 - 10.1109/TED.2018.2862139

DO - 10.1109/TED.2018.2862139

M3 - Article

AN - SCOPUS:85051660706

SN - 0018-9383

VL - 65

SP - 4716

EP - 4718

JO - IEEE Transactions on Electron Devices

JF - IEEE Transactions on Electron Devices

IS - 10

M1 - 8435959

ER -

Implementation of Convolutional Kernel Function Using 3-D TiO_x Resistive Switching Devices for Image Processing

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this