On Phase Information for Deep Neural Networks to Solve Full-Wave Nonlinear Inverse Scattering Problems

Xiao Min Pan; Bo Yue Song; Di Wu; Guohua Wei; Xin Qing Sheng

doi:10.1109/LAWP.2021.3100135

On Phase Information for Deep Neural Networks to Solve Full-Wave Nonlinear Inverse Scattering Problems

Xiao Min Pan^*, Bo Yue Song, Di Wu, Guohua Wei, Xin Qing Sheng

^*Corresponding author for this work

Beijing Institute of Technology

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

7 Citations (Scopus)

Abstract

The phase information's role in deep neural networks (DNNs) to solve the electromagnetic inverse scattering problems is investigated. The feedforward neutral network model with complex-valued (CV) data stream and the corresponding CV backpropagation training algorithm are proposed to realize CV convolutional neural networks. Numerical examples are carried out to demonstrate the phase information's role in DNNs in terms of generalization capability as well as the convergence speed in the training stage.

Original language	English
Pages (from-to)	1903-1907
Number of pages	5
Journal	IEEE Antennas and Wireless Propagation Letters
Volume	20
Issue number	10
DOIs	https://doi.org/10.1109/LAWP.2021.3100135
Publication status	Published - 1 Oct 2021

Keywords

Complex-valued (CV)
deep learning (DL)
electromagnetic inverse scattering
nonlinear

Access to Document

10.1109/LAWP.2021.3100135

Cite this

@article{51abd41d99d847cabdbc4d1fe9ee385d,

title = "On Phase Information for Deep Neural Networks to Solve Full-Wave Nonlinear Inverse Scattering Problems",

abstract = "The phase information's role in deep neural networks (DNNs) to solve the electromagnetic inverse scattering problems is investigated. The feedforward neutral network model with complex-valued (CV) data stream and the corresponding CV backpropagation training algorithm are proposed to realize CV convolutional neural networks. Numerical examples are carried out to demonstrate the phase information's role in DNNs in terms of generalization capability as well as the convergence speed in the training stage.",

keywords = "Complex-valued (CV), deep learning (DL), electromagnetic inverse scattering, nonlinear",

author = "Pan, {Xiao Min} and Song, {Bo Yue} and Di Wu and Guohua Wei and Sheng, {Xin Qing}",

note = "Publisher Copyright: {\textcopyright} 2002-2011 IEEE.",

year = "2021",

month = oct,

day = "1",

doi = "10.1109/LAWP.2021.3100135",

language = "English",

volume = "20",

pages = "1903--1907",

journal = "IEEE Antennas and Wireless Propagation Letters",

issn = "1536-1225",

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

number = "10",

}

TY - JOUR

T1 - On Phase Information for Deep Neural Networks to Solve Full-Wave Nonlinear Inverse Scattering Problems

AU - Pan, Xiao Min

AU - Song, Bo Yue

AU - Wu, Di

AU - Wei, Guohua

AU - Sheng, Xin Qing

PY - 2021/10/1

Y1 - 2021/10/1

N2 - The phase information's role in deep neural networks (DNNs) to solve the electromagnetic inverse scattering problems is investigated. The feedforward neutral network model with complex-valued (CV) data stream and the corresponding CV backpropagation training algorithm are proposed to realize CV convolutional neural networks. Numerical examples are carried out to demonstrate the phase information's role in DNNs in terms of generalization capability as well as the convergence speed in the training stage.

AB - The phase information's role in deep neural networks (DNNs) to solve the electromagnetic inverse scattering problems is investigated. The feedforward neutral network model with complex-valued (CV) data stream and the corresponding CV backpropagation training algorithm are proposed to realize CV convolutional neural networks. Numerical examples are carried out to demonstrate the phase information's role in DNNs in terms of generalization capability as well as the convergence speed in the training stage.

KW - Complex-valued (CV)

KW - deep learning (DL)

KW - electromagnetic inverse scattering

KW - nonlinear

UR - http://www.scopus.com/inward/record.url?scp=85112602236&partnerID=8YFLogxK

U2 - 10.1109/LAWP.2021.3100135

DO - 10.1109/LAWP.2021.3100135

M3 - Article

AN - SCOPUS:85112602236

SN - 1536-1225

VL - 20

SP - 1903

EP - 1907

JO - IEEE Antennas and Wireless Propagation Letters

JF - IEEE Antennas and Wireless Propagation Letters

IS - 10

ER -

On Phase Information for Deep Neural Networks to Solve Full-Wave Nonlinear Inverse Scattering Problems

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this