Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects

This study introduces a Self-Attention (SA) Generative Adversarial Network (GAN) framework that applies artificial intelligence techniques to microwave sensing for electromagnetic imaging. The approach involves illuminating anisotropic objects using Transverse Magnetic (TM) and Transverse Electric (...

Full description

Saved in:
Bibliographic Details
Main Authors: Chien-Ching Chiu, Po-Hsiang Chen, Yi-Hsun Chen, Hao Jiang
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/12/6723
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849467181981499392
author Chien-Ching Chiu
Po-Hsiang Chen
Yi-Hsun Chen
Hao Jiang
author_facet Chien-Ching Chiu
Po-Hsiang Chen
Yi-Hsun Chen
Hao Jiang
author_sort Chien-Ching Chiu
collection DOAJ
description This study introduces a Self-Attention (SA) Generative Adversarial Network (GAN) framework that applies artificial intelligence techniques to microwave sensing for electromagnetic imaging. The approach involves illuminating anisotropic objects using Transverse Magnetic (TM) and Transverse Electric (TE) electromagnetic waves, while sensing antennas collecting the scattered field data. To simplify the training process, a Back Propagation Scheme (BPS) is employed initially to calculate the preliminary permittivity distribution, which is then fed into the GAN with SA for image reconstruction. The proposed GAN with SA offers superior performance and higher resolution compared with GAN, along with enhanced generalization capability. The methodology consists of two main steps. First, TM waves are used to estimate the initial permittivity distribution along the z-direction using BPS. Second, TE waves estimate the x- and y-direction permittivity distribution. The estimated permittivity values are used as inputs to train the GAN with SA. In our study, we add 5% and 20% noise to compare the performance of the GAN with and without SA. Numerical results indicate that the GAN with SA demonstrates higher efficiency and resolution, as well as better generalization capability. Our innovation lies in the successful reconstruction of various uniaxial objects using a generator integrated with a self-attention mechanism, achieving reduced computational time and real-time imaging.
format Article
id doaj-art-1d1d0241449c408c90ffef84b5acb3fb
institution Kabale University
issn 2076-3417
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-1d1d0241449c408c90ffef84b5acb3fb2025-08-20T03:32:27ZengMDPI AGApplied Sciences2076-34172025-06-011512672310.3390/app15126723Self-Attention GAN for Electromagnetic Imaging of Uniaxial ObjectsChien-Ching Chiu0Po-Hsiang Chen1Yi-Hsun Chen2Hao Jiang3Department of Electrical and Computer Engineering, Tamkang University, Tamsui 251301, TaiwanDepartment of Electrical and Computer Engineering, Tamkang University, Tamsui 251301, TaiwanDepartment of Electrical and Computer Engineering, Tamkang University, Tamsui 251301, TaiwanSchool of Engineering, San Francisco State University, San Francisco, CA 94117-1080, USAThis study introduces a Self-Attention (SA) Generative Adversarial Network (GAN) framework that applies artificial intelligence techniques to microwave sensing for electromagnetic imaging. The approach involves illuminating anisotropic objects using Transverse Magnetic (TM) and Transverse Electric (TE) electromagnetic waves, while sensing antennas collecting the scattered field data. To simplify the training process, a Back Propagation Scheme (BPS) is employed initially to calculate the preliminary permittivity distribution, which is then fed into the GAN with SA for image reconstruction. The proposed GAN with SA offers superior performance and higher resolution compared with GAN, along with enhanced generalization capability. The methodology consists of two main steps. First, TM waves are used to estimate the initial permittivity distribution along the z-direction using BPS. Second, TE waves estimate the x- and y-direction permittivity distribution. The estimated permittivity values are used as inputs to train the GAN with SA. In our study, we add 5% and 20% noise to compare the performance of the GAN with and without SA. Numerical results indicate that the GAN with SA demonstrates higher efficiency and resolution, as well as better generalization capability. Our innovation lies in the successful reconstruction of various uniaxial objects using a generator integrated with a self-attention mechanism, achieving reduced computational time and real-time imaging.https://www.mdpi.com/2076-3417/15/12/6723electromagnetic imaginguniaxial objectsU-Netdeep learninginverse scatteringgenerative adversarial network
spellingShingle Chien-Ching Chiu
Po-Hsiang Chen
Yi-Hsun Chen
Hao Jiang
Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
Applied Sciences
electromagnetic imaging
uniaxial objects
U-Net
deep learning
inverse scattering
generative adversarial network
title Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
title_full Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
title_fullStr Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
title_full_unstemmed Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
title_short Self-Attention GAN for Electromagnetic Imaging of Uniaxial Objects
title_sort self attention gan for electromagnetic imaging of uniaxial objects
topic electromagnetic imaging
uniaxial objects
U-Net
deep learning
inverse scattering
generative adversarial network
url https://www.mdpi.com/2076-3417/15/12/6723
work_keys_str_mv AT chienchingchiu selfattentionganforelectromagneticimagingofuniaxialobjects
AT pohsiangchen selfattentionganforelectromagneticimagingofuniaxialobjects
AT yihsunchen selfattentionganforelectromagneticimagingofuniaxialobjects
AT haojiang selfattentionganforelectromagneticimagingofuniaxialobjects