xudong chen

Physics-Assisted Machine Learning for Solving Electromagnetic Inverse Problems

This talk applies machine learning (ML) to solve electromagnetic inverse problems, which in fact cover various reconstruction problems (also known as parameter identification problems), optimal design problems (also known as inverse designs or optimal synthesis), and classification problems. As examples, this talk demonstrates how we apply ML to solve radar imaging and classification problems. Solving wave imaging problems using ML has attracted researchers’ interests in recent years. Nevertheless, most existing works directly adopt ML as a black box. In fact, researchers have gained, over several decades, much insightful domain knowledge on wave physics and in addition some of these physical laws present well-known mathematical properties (even analytical formulas), which do not need to be learnt by training with a lot of data. This talk demonstrates that it is of paramount importance to profitably combine ML with the available knowledge on underlying wave physics. Besides imaging, another important application of wave sensing is target classification. This talk proposes a high-accuracy and efficient classification method with ML on frequency-modulated-continuous-wave (FMCW) radar. Instead of directly choosing measured data as the input of neural network, we start from the first principle of wave physics to design low-dimensional input. The proposed classifier is applied to automotive radar system, where road targets are to be classified into five categories, including pedestrian, bike, sedan, truck/bus, and other static objects. The proposed physics-assisted classifier is tested with real-world data obtained from 77-GHz FMCW radars and it turns out to competitive among the state-of-the-art methods in automotive radar application. Additionally, in line with the theme of the Forum, we tackle certain antenna design problems from the perspective of scattering and discuss how physics of radiation and scattering can be combined with ML. Finally, some experiences gained by solving various inverse problems are summarized.

Xudong CHEN received the B.S. and M.S. degrees from Zhejiang University, China and the Ph.D. degree from the Massachusetts Institute of Technology, USA. Since 2005, he has been with the National University of Singapore, Singapore, where he is currently a Professor. He has published 170 journal papers on inverse scattering problems, material parameter retrieval, microscopy, optical encryption, and physics-assisted machine learning. He has authored the book Computational Methods for Electromagnetic Inverse Scattering (Wiley-IEEE, 2018). His research interests include mainly electromagnetic wave theories and applications, with a focus on inverse problems and computational imaging. Dr. Chen is a Fellow of IEEE and Fellow of the Electromagnetics Academy. He was a recipient of the Young Scientist Award by the Union Radio Scientifique Internationale in 2010 and a recipient of the Ulrich L. Rohde Innovative Conference Paper Award at ICCEM 2019 conference. He is currently a Deputy Editor-in-Chief (D-EiC) of IEEE Transactions on Geoscience and Remote Sensing. He was an Associate Editor (AE) of IEEE J-ERM, IEEE T-MTT, and IEEE T-GRS. He has been members of organizing committees of more than 10 conferences, serving as General Chair, Technical Program Committee (TPC) Chair, Award Committee Chair, etc. He was the Chair of IEEE Singapore Microwave Theory and Techniques (MTT)/Antennas and Propagation (AP) Joint Chapter in 2018.