Unrolled Convolutional Neural Network for Full-Wave Inverse Scattering

被引：16

作者：

Zhang, Yarui ^{[1
]}

Lambert, Marc ^{[1
]}

Fraysse, Aurelia ^{[2
]}

Lesselier, Dominique ^{[2
]}

机构：

[1] Univ Paris Saclay, Lab Genie Elect & Elect Paris, CentraleSupelec, CNRS, F-91192 Gif Sur Yvette, France

[2] Univ Paris Saclay, Lab Signaux & Syst, CNRS, CentraleSupelec, F-91190 Gif Sur Yvette, France

来源：

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION | 2023年 / 71卷 / 01期

关键词：

Contrast source inversion (CSI) method; deep learning; inverse scattering; unrolled method;

D O I：

10.1109/TAP.2022.3216999

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

An unrolled deep learning scheme for solving full-wave nonlinear inverse scattering problems (ISPs) is proposed. Inspired by the so- called unrolled method, an iterative neural network structure combining the contrast source inversion (CSI) method and residual network (ResNet) is designed. By embedding the CSI iterations into the deep learning model, the domain knowledge is well incorporated into the learning process. Thorough numerical tests are carried out to evaluate the performance, stability, robustness, and reliability of the proposed approach. Comparisons with the widely used U-net structure and CSI exhibit the advantage of the proposed approach.

引用

页码：947 / 956

页数：10

共 50 条

[21] Full-wave Electromagnetic Scattering Simulation of Plunging Breaking Waves
Wu, Biyi
Wang, Jiajing
Sheng, Xinqing
2018 12TH INTERNATIONAL SYMPOSIUM ON ANTENNAS, PROPAGATION AND ELECTROMAGNETIC THEORY (ISAPE), 2018,
[22] Performance Analysis and Dynamic Evolution of Deep Convolutional Neural Network for Electromagnetic Inverse Scattering
Li, Lianlin
Wang, Long Gang
Teixeira, Fernando L.
IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, 2019, 18 (11): : 2259 - 2263
[23] Solving the inverse problem in OCT using full-wave adjoint models
Macdonald, Callum M.
Arridge, Simon R.
Munro, Peter R. T.
2019 IEEE PHOTONICS CONFERENCE (IPC), 2019,
[24] Reducing a Number of Full-Wave Analyses in RBF Neural Network Optimization of Complex Microwave Structures
Murphy, Ethan K.
Yakovlev, Vadim V.
2009 IEEE/MTT-S INTERNATIONAL MICROWAVE SYMPOSIUM, VOLS 1-3, 2009, : 1253 - 1256
[25] Full-Wave and Approximate Solutions of Large Electromagnetic Scattering Problems
Hidayetoglu, Mert
Gurel, Levent
2015 IEEE INTERNATIONAL SYMPOSIUM ON ANTENNAS AND PROPAGATION & USNC/URSI NATIONAL RADIO SCIENCE MEETING, 2015, : 566 - 567
[26] Solving Inverse Scattering Problem with a Crack in Inhomogeneous Medium Based on a Convolutional Neural Network
Yin, Weishi
Yang, Ziyuan
Meng, Pinchao
SYMMETRY-BASEL, 2023, 15 (01):
[27] ON THE FULL-WAVE DIPOLE
STUMPFF, G
IEEE ANTENNAS AND PROPAGATION MAGAZINE, 1995, 37 (05) : 51 - 52
[28] On Obtaining Initial Approximation for the Full Wave Inversion Problem Using Convolutional Neural Network
Petrov, I. B.
Stankevich, A. S.
Vasyukov, A. V.
DOKLADY MATHEMATICS, 2023, 108 (01) : 291 - 296
[29] On Obtaining Initial Approximation for the Full Wave Inversion Problem Using Convolutional Neural Network
I. B. Petrov
A. S. Stankevich
A. V. Vasyukov
Doklady Mathematics, 2023, 108 : 291 - 296
[30] Network Methods for Full-Wave Modeling of Stochastic Electromagnetic Fields
Russer, J. A.
Haider, M.
Russer, P.
2017 INTERNATIONAL CONFERENCE ON ELECTROMAGNETICS IN ADVANCED APPLICATIONS (ICEAA), 2017, : 1494 - 1497

← 1 2 3 4 5 →