A new kind of electromagnetic wave, combining the properties of propagating and evanescent waves, is discovered theoretically in a paper by E. Narimanov, published in this issue of Advanced Photonics. The commentary provides context for Narimov’s paper.

Optical computing provides unique opportunities in terms of parallelization, scalability, power efficiency, and computational speed and has attracted major interest for machine learning. Diffractive deep neural networks have been introduced earlier as an optical machine learning framework that uses task-specific diffractive surfaces designed by deep learning to all-optically perform inference, achieving promising performance for object classification and imaging. We demonstrate systematic improvements in diffractive optical neural networks, based on a differential measurement technique that mitigates the strict nonnegativity constraint of light intensity. In this differential detection scheme, each class is assigned to a separate pair of detectors, behind a diffractive optical network, and the class inference is made by maximizing the normalized signal difference between the photodetector pairs. Using this differential detection scheme, involving 10 photodetector pairs behind 5 diffractive layers with a total of 0.2 million neurons, we numerically achieved blind testing accuracies of 98.54%, 90.54%, and 48.51% for MNIST, Fashion-MNIST, and grayscale CIFAR-10 datasets, respectively. Moreover, by utilizing the inherent parallelization capability of optical systems, we reduced the cross-talk and optical signal coupling between the positive and negative detectors of each class by dividing the optical path into two jointly trained diffractive neural networks that work in parallel. We further made use of this parallelization approach and divided individual classes in a target dataset among multiple jointly trained diffractive neural networks. Using this class-specific differential detection in jointly optimized diffractive neural networks that operate in parallel, our simulations achieved blind testing accuracies of 98.52%, 91.48%, and 50.82% for MNIST, Fashion-MNIST, and grayscale CIFAR-10 datasets, respectively, coming close to the performance of some of the earlier generations of all-electronic deep neural networks, e.g., LeNet, which achieves classification accuracies of 98.77%, 90.27%, and 55.21% corresponding to the same datasets, respectively. In addition to these jointly optimized diffractive neural networks, we also independently optimized multiple diffractive networks and utilized them in a way that is similar to ensemble methods practiced in machine learning; using 3 independently optimized differential diffractive neural networks that optically project their light onto a common output/detector plane, we numerically achieved blind testing accuracies of 98.59%, 91.06%, and 51.44% for MNIST, Fashion-MNIST, and grayscale CIFAR-10 datasets, respectively. Through these systematic advances in designing diffractive neural networks, the reported classification accuracies set the state of the art for all-optical neural network design. The presented framework might be useful to bring optical neural network-based low power solutions for various machine learning applications and help us design new computational cameras that are task-specific.

Synchronization is of importance in both fundamental and applied physics, but its demonstration at the micro/nanoscale is mainly limited to low-frequency oscillations such as mechanical resonators. We report the synchronization of two coupled optical microresonators, in which the high-frequency resonances in the optical domain are aligned with reduced noise. It is found that two types of synchronization regimes emerge with either the first- or second-order transition, both presenting a process of spontaneous symmetry breaking. In the second-order regime, the synchronization happens with an invariant topological character number and a larger detuning than that of the first-order case. Furthermore, an unconventional hysteresis behavior is revealed for a time-dependent coupling strength, breaking the static limitation and the temporal reciprocity. The synchronization of optical microresonators offers great potential in reconfigurable simulations of many-body physics and scalable photonic devices on a chip.

We show that dielectric waveguides formed by materials with strong optical anisotropy support electromagnetic waves that combine the properties of propagating and evanescent fields. These “ghost waves” are created in tangent bifurcations that “annihilate” pairs of positive- and negative-index modes and represent the optical analogue of the “ghost orbits” in the quantum theory of nonintegrable dynamical systems. Ghost waves can be resonantly coupled to the incident evanescent field, which then grows exponentially through the anisotropic media—as in the case of negative index materials. As ghost waves are supported by transparent dielectric media, the proposed approach to electromagnetic field enhancement is free from the “curse” of material loss that is inherent to conventional negative index composites.

Bose–Einstein condensate (BEC) exhibits a variety of fascinating and unexpected macroscopic phenomena, and has attracted sustained attention in recent years—particularly in the field of solitons and associated nonlinear phenomena. Meanwhile, optical lattices have emerged as a versatile toolbox for understanding the properties and controlling the dynamics of BEC, among which the realization of bright gap solitons is an iconic result. However, the dark gap solitons are still experimentally unproven, and their properties in more than one dimension remain unknown. In light of this, we describe, numerically and theoretically, the formation and stability properties of gap-type dark localized modes in the context of ultracold atoms trapped in optical lattices. Two kinds of stable dark localized modes—gap solitons and soliton clusters—are predicted in both the one- and two-dimensional geometries. The vortical counterparts of both modes are also constructed in two dimensions. A unique feature is the existence of a nonlinear Bloch-wave background on which all above gap modes are situated. By employing linear-stability analysis and direct simulations, stability regions of the predicted modes are obtained. Our results offer the possibility of observing dark gap localized structures with cutting-edge techniques in ultracold atoms experiments and beyond, including in optics with photonic crystals and lattices.

Entanglement distribution between distant parties is one of the most important and challenging tasks in quantum communication. Distribution of photonic entangled states using optical fiber links is a fundamental building block toward quantum networks. Among the different degrees of freedom, orbital angular momentum (OAM) is one of the most promising due to its natural capability to encode high dimensional quantum states. We experimentally demonstrate fiber distribution of hybrid polarization-vector vortex entangled photon pairs. To this end, we exploit a recently developed air-core fiber that supports OAM modes. High fidelity distribution of the entangled states is demonstrated by performing quantum state tomography in the polarization-OAM Hilbert space after fiber propagation and by violations of Bell inequalities and multipartite entanglement tests. The results open new scenarios for quantum applications where correlated complex states can be transmitted by exploiting the vectorial nature of light.