Back-propagation optimization and multi-valued artificial neural networks for highly vivid structural color filter metasurfaces.

We introduce a novel technique for designing color filter metasurfaces using a data-driven approach based on deep learning. Our innovative approach employs inverse design principles to identify highly efficient designs that outperform all the configurations in the dataset, which consists of 585 distinct geometries solely. By combining Multi-Valued Artificial Neural Networks and back-propagation optimization, we overcome the limitations of previous approaches, such as poor performance due to extrapolation and undesired local minima.

Design of a novel hybrid multimode interferometer operating with both TE and TM polarizations for sensing applications.

A novel hybrid multimode interferometer for sensing applications operating with both TE and TM polarizations simultaneously is proposed and numerically demonstrated. The simulations were performed assuming an operating wavelength of 633 nm with the goal of future use as a biosensor, but its applications extend beyond that area and could be adapted for any wavelength or application of interest. By designing the mutimode waveguide core with a low aspect ratio, the confinement characteristics of TE modes and TM modes become very distinct and their interaction with the sample in the sensing area becomes very different as well, resulting in high device sensitivity.

Quasi-dark resonances with antiferromagnetic order in silicon metasurfaces.

Quasi-dark resonances exhibiting antiferromagnetic order are theoretically investigated in a near-infrared metasurface composed of square slotted rings etched in a thin silicon layer on glass substrate. Access to the quasi-dark mode is achieved by reducing the symmetry of the metasurface according to the findings of a detailed group theory analysis. A thorough finite-element study reveals the key optical properties of the antiferromagnetic order quasi-dark mode, namely resonant wavelengths, quality factors, angular dispersion, and its robustness against optical extinction losses.

Spiral SAR Imaging with Fast Factorized Back-Projection: A Phase Error Analysis.

This paper presents a fast factorized back-projection (FFBP) algorithm that can satisfactorily process real P-band synthetic aperture radar (SAR) data collected from a spiral flight pattern performed by a drone-borne SAR system. Choosing the best setup when processing SAR data with an FFBP algorithm is not so straightforward, so predicting how this choice will affect the quality of the output image is valuable information. This paper provides a statistical phase error analysis to validate the hypothesis that the phase error standard deviation can be predicted by geometric parameters specified at the start of processing.