Deep inverse methods in electromagnetic metamaterials

Willie J. Padilla

doi:10.1117/12.3014726

12 March 2024 Deep inverse methods in electromagnetic metamaterials

Willie J. Padilla

Proceedings Volume PC12885, Terahertz, RF, Millimeter, and Submillimeter-Wave Technology and Applications XVII; PC128850N (2024) https://doi.org/10.1117/12.3014726
Event: SPIE OPTO, 2024, San Francisco, California, United States

Abstract

Deep learning (DL) has enabled the development of deep inverse models (DIMs) to solve inverse problems in artificial electromagnetic materials (AEMs). DIMs often outperform conventional optimization approaches, but their performance has not been thoroughly compared. We evaluated eight state-of-the-art DIMs on three unique AEM design problems, quantitatively comparing their solution time and accuracy. We found that modern DIMs can be decomposed into independent modules, and that interchanging these modules can create novel higher performing DIMs. We taxonomized the unique modules of current state-of-the-art DIMs into three categories: initializers, filters, and optimizers. We conclude by discussing some important outstanding issues of deep inverse design of AEMs, and presenting an outlook of this exciting field.

Conference Presentation

Citation Download Citation

Willie J. Padilla "Deep inverse methods in electromagnetic metamaterials", Proc. SPIE PC12885, Terahertz, RF, Millimeter, and Submillimeter-Wave Technology and Applications XVII, PC128850N (12 March 2024); https://doi.org/10.1117/12.3014726

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available