@inproceedings{4ac8ebe4769348e7a6d298f6442e1c45,
title = "Multiresolution Physics-Informed Deep Learning to Solve Electromagnetic Inverse Scattering Problems",
abstract = "A multiresolution physics-informed deep learning (DL) method is proposed to solve electromagnetic inverse scattering problems (ISPs) by embedding the objective function of the traditional full-wave solver into the loss function of the physics-informed neural network (PINN) within a multi-focusing scheme. Numerical experiments are carried out to demonstrate the proposed DL-based method outperforms the traditional full-wave solver in terms of inversion accuracy.",
keywords = "Physics-informed neural networks (PINNs), electromagnetic inverse scattering, iterative multi-scaling",
author = "Wang, \{Ji Yuan\} and Liang Zhang and Wang, \{Yun Chuan\} and Zhao, \{Cheng Tao\} and Lou, \{Xin Yue\} and Pan, \{Xiao Min\}",
note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 8th IEEE International Symposium on Electromagnetic Compatibility, ISEMC 2025 ; Conference date: 10-10-2025 Through 12-10-2025",
year = "2025",
doi = "10.1109/ISEMC68048.2025.11292597",
language = "English",
series = "ISEMC 2025 - 8th International Symposium on Electromagnetic Compatibility, Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "ISEMC 2025 - 8th International Symposium on Electromagnetic Compatibility, Proceedings",
address = "United States",
}