Integrated diffractive deep neural networks for near-infrared and visible wavelengths
Abstract
Diffractive deep neural networks (DDNNs) are free-space optical image processors that can perform image classifications. Integrated DDNNs can be directly fabricated on detectors with two-photon polymerization (TPP). This publication discusses some of the numerical and fabrication limitations that govern the design of integrated DDNNs for near-infrared (NIR) and visible wavelengths. Numerical optimization and simulation results are presented for an example geometry that can be fabricated with TPP printing.
Fields of Study & Subjects
Neural Networks and Reservoir Computing
Metamaterials and Metasurfaces Applications
Random lasers and scattering media
Publication details
- Journal / source
- AI and Optical Data Sciences
- Pages
- 40–40
- Publication date
- 4 March 2026
- Language
- EN