Phase imaging with an untrained neural network

Most of the neural networks proposed so far for computational imaging (CI) in optics employ a supervised training strategy, and thus need a large training set to optimize their weights and biases. Setting aside the requirements of environmental and system stability during many hours of data acquisition, in many practical applications, it is unlikely to be possible to obtain sufficient numbers of ground-truth images for training. Here, we propose to overcome this limitation by incorporating into a conventional deep neural network a complete physical model that represents the process of image formation. The most significant advantage of the resulting physics-enhanced deep neural network (PhysenNet) is that it can be used without training beforehand, thus eliminating the need for tens of thousands of labeled data. We take single-beam phase imaging as an example for demonstration. We experimentally show that one needs only to feed PhysenNet a single diffraction pattern of a phase object, and it can automatically optimize the network and eventually produce the object phase through the interplay between the neural network and the physical model. This opens up a new paradigm of neural network design, in which the concept of incorporating a physical model into a neural network can be generalized to solve many other CI problems.

Nanowire Length, Density, and Crystalline Quality Retrieved from a Single Optical Spectrum

We propose spectral domain attenuated reflectometry (SDAR) for fast characterization of nanomaterial growth. The method is demonstrated here for zinc oxide (ZnO) nanowires (NWs) which are grown vertically in random forest fashion showing that it is not limited to well-ordered NWs. We show how SDAR can provide, on the basis of a single measured spectrum, simultaneous information on nanowire length, nanowire density (through nanowire/air filling ratio), and crystalline quality (through band gap). The robustness of the proposed method is assessed first through comparison with information obtained from SEM and XRD taken as reference. In SDAR, the process for fast extraction of NW thickness and filling ratio values  makes use of the interference pattern contrast and the spectral periodicity in the reflection response which involve a best fit of the measured spectra with simple theoretical modeling based on the effective medium approach, achieved with a mean square error down to 0.1%. The results also suggest the existence of either 2 or 3 layers of different effective refractive index, hence providing insight on possible growth mechanisms.

Quantum expander for gravitational-wave observatories

The quantum uncertainty of laser light limits the sensitivity of gravitational-wave observatories. Over the past 30 years, techniques for squeezing the quantum uncertainty, as well as for enhancing gravitational-wave signals with optical resonators have been invented. Resonators, however, have finite linewidths, and the high signal frequencies that are produced during the highly scientifically interesting ring-down of astrophysical compact-binary mergers still cannot be resolved. Here, we propose a purely optical approach for expanding the detection bandwidth. It uses quantum uncertainty squeezing inside one of the optical resonators, compensating for the finite resonators' linewidths while keeping the low-frequency sensitivity unchanged. This quantum expander is intended to enhance the sensitivity of future gravitational-wave detectors, and we suggest the use of this new tool in other cavity-enhanced metrological experiments.

In situ characterization of nanowire dimensions and growth dynamics by optical reflectance

Optical reflectometry is commonly used as an accurate and noninvasive characterization tool when growing planar semiconductor layers. However, thin-film analysis schemes cannot be directly applied to nanowire systems due to their complex optical response. Here, we report on reliable in situ characterization of nanowire growth with high accuracy using optical reflectance spectra for analysis. The method makes it possible to determine the nanowire length, diameter, and growth rate in situ in real time with high resolution. We demonstrate the method's versatility by using the optical reflectance data for determining nanowire dimensions on both particle-assisted and selective-area grown nanowires. To indicate the full potential of in situ characterization of nanowire synthesis we evaluate the growth dynamics of InP nanowires in the presence of the p-type dopant precursor diethylzinc. We observe that the growth rate is strongly affected by the diethylzinc. At low diethylzinc flows, the growth rate decreases monotonously while higher flows lead to an initially increasing growth rate. From these in situ characterization data, we conclude that the surface migration length of adatom species is affected strongly by the addition of diethylzinc. We believe that this characterization method will become a standard tool for in situ growth monitoring and aid in elucidating the complex growth dynamics often exhibited during nanowire growth.

Information multiplexing in ptychography

We show for the first time that ptychography (a form of lensless diffractive imaging) can recover the spectral response of an object through simultaneous reconstruction of multiple images that represent the object's response to a particular mode present in the illumination. We solve the phase problem for each mode independently, even though the intensity arriving at every detector pixel is an incoherent superposition of several uncorrelated diffracted waves. Until recently, the addition of incoherent modes has been seen as a nuisance in diffractive imaging: here we show that not only can the difficulties they pose be removed, but that they can also be used to discover much more information about the object. If the illumination function is also mode-specific, we show that we can also solve simultaneously for a multiplicity of such illumination modes. The work opens exciting possibilities for information multiplexing in ptychography over all visible, X-ray and electron wavelengths.