Related Experiment Video
Updated: Aug 25, 2025
09:33
Demonstration of Equal-Intensity Beam Generation by Dielectric Metasurfaces
Published on: June 7, 2019
6.3K
Deep-Learning-Empowered Holographic Metasurface with Simultaneously Customized Phase and Amplitude
Ruichao Zhu1, Jiafu Wang1, Xinmin Fu1
1Shaanxi Key Laboratory of Artificially-Structured Functional Materials and Devices, Air Force Engineering University, Xi'an, Shaanxi710051, China.
ACS Applied Materials & Interfaces
|October 17, 2022
Summary
This study introduces an AI-driven inverse design for meta-atoms, enabling independent control of amplitude and phase for electromagnetic waves. This significantly speeds up the design of high-resolution holographic metasurfaces.
Area of Science:
- Electromagnetic Wave Manipulation
- Metasurface Design
- Artificial Materials
Background:
- Metasurfaces offer advanced control over electromagnetic (EM) waves by independently modulating amplitude and phase.
- Simultaneous amplitude and phase control enables higher resolution in holographic applications.
- Current design processes for such metasurfaces are complex and time-consuming.
Purpose of the Study:
- To develop an inverse design method for meta-atoms capable of simultaneously and independently controlling transmitted wave amplitude and phase.
- To significantly enhance the design efficiency of holographic metasurfaces.
- To eliminate parameter coupling in meta-atom design.
Main Methods:
- Utilized a customized deep ResNet for inverse design of meta-atoms.
- Trained the network to tailor amplitude and phase responses while decoupling parameters.
- Designed and fabricated two holographic metasurfaces using the trained network without parameter sweeping.
Main Results:
- Successfully designed meta-atoms with independent amplitude and phase control.
- Achieved high-resolution holography demonstrated through both simulated and measured results.
- Significantly enhanced design efficiency compared to traditional methods.
Conclusions:
- The proposed AI-driven inverse design method is reliable for creating high-resolution holographic metasurfaces.
- This approach drastically reduces design time and complexity.
- The methodology holds potential for intelligent design across various artificial materials and surfaces.
