Big-Volume SliceGAN for Improving a Synthetic 3D Microstructure Image of Additive-Manufactured TYPE 316L Steel

Keiya Sugiura; Toshio Ogawa; Yoshitaka Adachi; Fei Sun; Asuka Suzuki; Akinori Yamanaka; Nobuo Nakada; Takuya Ishimoto; Takayoshi Nakano; Yuichiro Koizumi

doi:10.3390/jimaging9050090

Big-Volume SliceGAN for Improving a Synthetic 3D Microstructure Image of Additive-Manufactured TYPE 316L Steel

Keiya Sugiura, Toshio Ogawa, Yoshitaka Adachi^*, Fei Sun, Asuka Suzuki, Akinori Yamanaka, Nobuo Nakada, Takuya Ishimoto, Takayoshi Nakano, Yuichiro Koizumi

^*Corresponding author for this work

Center for Aluminum and advanced Materials Research and International Collaboration, Institute of Light Metals

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

7 Scopus citations

Abstract

A modified SliceGAN architecture was proposed to generate a high-quality synthetic three-dimensional (3D) microstructure image of TYPE 316L material manufactured through additive methods. The quality of the resulting 3D image was evaluated using an auto-correlation function, and it was discovered that maintaining a high resolution while doubling the training image size was crucial in creating a more realistic synthetic 3D image. To meet this requirement, modified 3D image generator and critic architecture was developed within the SliceGAN framework.

Original language	English
Article number	90
Journal	Journal of Imaging
Volume	9
Issue number	5
DOIs	https://doi.org/10.3390/jimaging9050090
State	Published - 2023/05

Keywords

SliceGAN
additive manufacturing
autocorrelation function
generative adversarial network
synthetic 3D image

ASJC Scopus subject areas

Radiology Nuclear Medicine and imaging
Computer Vision and Pattern Recognition
Computer Graphics and Computer-Aided Design
Electrical and Electronic Engineering

Access to Document

10.3390/jimaging9050090

Cite this

@article{ce7202377d694c3da8061bdf5fb44976,

title = "Big-Volume SliceGAN for Improving a Synthetic 3D Microstructure Image of Additive-Manufactured TYPE 316L Steel",

abstract = "A modified SliceGAN architecture was proposed to generate a high-quality synthetic three-dimensional (3D) microstructure image of TYPE 316L material manufactured through additive methods. The quality of the resulting 3D image was evaluated using an auto-correlation function, and it was discovered that maintaining a high resolution while doubling the training image size was crucial in creating a more realistic synthetic 3D image. To meet this requirement, modified 3D image generator and critic architecture was developed within the SliceGAN framework.",

keywords = "SliceGAN, additive manufacturing, autocorrelation function, generative adversarial network, synthetic 3D image",

author = "Keiya Sugiura and Toshio Ogawa and Yoshitaka Adachi and Fei Sun and Asuka Suzuki and Akinori Yamanaka and Nobuo Nakada and Takuya Ishimoto and Takayoshi Nakano and Yuichiro Koizumi",

note = "Publisher Copyright: {\textcopyright} 2023 by the authors.",

year = "2023",

month = may,

doi = "10.3390/jimaging9050090",

language = "英語",

volume = "9",

journal = "Journal of Imaging",

issn = "2313-433X",

publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",

number = "5",

}

TY - JOUR

T1 - Big-Volume SliceGAN for Improving a Synthetic 3D Microstructure Image of Additive-Manufactured TYPE 316L Steel

AU - Sugiura, Keiya

AU - Ogawa, Toshio

AU - Adachi, Yoshitaka

AU - Sun, Fei

AU - Suzuki, Asuka

AU - Yamanaka, Akinori

AU - Nakada, Nobuo

AU - Ishimoto, Takuya

AU - Nakano, Takayoshi

AU - Koizumi, Yuichiro

PY - 2023/5

Y1 - 2023/5

N2 - A modified SliceGAN architecture was proposed to generate a high-quality synthetic three-dimensional (3D) microstructure image of TYPE 316L material manufactured through additive methods. The quality of the resulting 3D image was evaluated using an auto-correlation function, and it was discovered that maintaining a high resolution while doubling the training image size was crucial in creating a more realistic synthetic 3D image. To meet this requirement, modified 3D image generator and critic architecture was developed within the SliceGAN framework.

AB - A modified SliceGAN architecture was proposed to generate a high-quality synthetic three-dimensional (3D) microstructure image of TYPE 316L material manufactured through additive methods. The quality of the resulting 3D image was evaluated using an auto-correlation function, and it was discovered that maintaining a high resolution while doubling the training image size was crucial in creating a more realistic synthetic 3D image. To meet this requirement, modified 3D image generator and critic architecture was developed within the SliceGAN framework.

KW - SliceGAN

KW - additive manufacturing

KW - autocorrelation function

KW - generative adversarial network

KW - synthetic 3D image

UR - http://www.scopus.com/inward/record.url?scp=85160303848&partnerID=8YFLogxK

U2 - 10.3390/jimaging9050090

DO - 10.3390/jimaging9050090

M3 - 学術論文

C2 - 37233310

AN - SCOPUS:85160303848

SN - 2313-433X

VL - 9

JO - Journal of Imaging

JF - Journal of Imaging

IS - 5

M1 - 90

ER -

Big-Volume SliceGAN for Improving a Synthetic 3D Microstructure Image of Additive-Manufactured TYPE 316L Steel

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this