We aim to address one of the fundamental limitations of machine learning (ML): its reliance on extensive training datasets by incorporating physics-based intuition and Maxwell-equation-based constraints into ML process. We show that physics-guided networks require significantly smaller datasets, enable learning outside the original training data, and provide improved prediction accuracy and physics consistency. The proposed approaches are illustrated on examples of photonic composites, from photonic crystals to hyperbolic metamaterials.
Pixel size in cameras and other refractive imaging devices is typically limited by the free-space diffraction. However, a vast majority of semiconductor-based detectors are based on materials with substantially high refractive index. We demonstrate that diffractive optics can be used to take advantage of this high refractive index to reduce effective pixel size of the sensors below free-space diffraction limit. At the same time, diffractive systems encode both amplitude and phase information about the incoming beam into multiple pixels, offering the platform for noise-tolerant imaging with dynamical refocusing. We explore the opportunities opened by high index diffractive optics to reduce sensor size and increase signalto- noise ratio of imaging structures.
Optical characterization of subwavelength objects is important for biology, nanotechnology, chemistry, and materials science. Unfortunately, the information about interaction of an isolated subwavelength object with light is contained in evanescent waves that exponentially decay away from the source. Numerous techniques have been proposed to access or restore this information. In interscale mixing microscopy (IMM), a diffraction grating positioned in the near field proximity of the object is used to convert the originally-evanescent waves into propagating modes that can be detected with far-field measurements. However, far-field signal needs to be post-processed to un-couple the diffraction-limited and subwavelength responses. Several techniques, based on multiple measurements, have been previously proposed. Here, we show that with simple Fourier-transform based post processing can be used to characterize position, and optical size of the object based on a single measurement. To verify the proposed formalism, three finite diffraction gratings were fabricated. Two of these gratings contained pre-engineered “defects” that played the role of “unknown objects”, while the remaining grating was used as a reference. We demonstrate that we can identify the position and size of ~wavelength/10 object with far-field characterization. The same measurement provides a platform to analyze optical spectrum of the object. Although demonstrated in this work on example of 1D grating, IMM can be extended to 2D subwavelength imaging
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
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.