Following a recent theoretical suggestion and microwave experiments, we fabricate photonic metamaterials composed of pairs of twisted gold crosses using two successive electron-beam-lithography steps and intermediate planarization via a spin-on dielectric. The resulting two effective resonances of the coupled system lie in the 1-2 microm wavelength regime and exhibit pronounced circular dichroism, while the circular polarization conversion is very small. In between the two resonances, we find a fairly broad spectral regime with strong optical activity, i.e., with a pure rotation of incident linear polarization. The measured optical transmittance spectra agree well with theory.

Abstract. In this work we present a novel approach for single depth map super-resolution. Modern consumer depth sensors, especially Timeof-Flight sensors, produce dense depth measurements, but are affected by noise and have a low lateral resolution. We propose a method that combines the benefits of recent advances in machine learning based single image super-resolution, i.e. deep convolutional networks, with a variational method to recover accurate high-resolution depth maps. In particular, we integrate a variational method that models the piecewise affine structures apparent in depth data via an anisotropic total generalized variation regularization term on top of a deep network. We call our method ATGV-Net and train it end-to-end by unrolling the optimization procedure of the variational method. To train deep networks, a large corpus of training data with accurate ground-truth is required. We demonstrate that it is feasible to train our method solely on synthetic data that we generate in large quantities for this task. Our evaluations show that we achieve state-of-the-art results on three different benchmarks, as well as on a challenging Time-of-Flight dataset, all without utilizing an additional intensity image as guidance.

We study arrays of silver split-ring resonators operating at around 1.5-µm wavelength coupled to an MBE-grown single 12.7-nm thin InGaAs quantum well separated only 4.8 nm from the wafer surface. The samples are held at liquid-helium temperature and are pumped by intense femtosecond optical pulses at 0.81-µm center wavelength in a pump-probe geometry. We observe much larger relative transmittance changes (up to about 8%) on the split-ring-resonator arrays as compared to the bare quantum well (not more than 1-2%). We also observe a much more rapid temporal decay component of the differential transmittance signal of 15 ps for the case of split-ring resonators coupled to the quantum well compared to the case of the bare quantum well, where we find about 0.7 ns. These observations are ascribed to the evanescent coupling of the split-ring resonators to the quantum-well gain. All experimental results are compared with a recently introduced analytical toy model that accounts for this evanescent coupling, leading to excellent overall qualitative agreement.

Abstract:We propose, solve, and discuss a simple model for a metamaterial incorporating optical gain: A single bosonic resonance is coupled to a fermionic (inverted) two-level-system resonance via local-field interactions. For given steady-state inversion, this model can be solved analytically, revealing a rich variety of (Fano) absorption/gain lineshapes. We also give an analytic expression for the fixed inversion resulting from gain pinning under steady-state conditions. Furthermore, the dynamic response of the "lasing SPASER", i.e., its relaxation oscillations, can be obtained by simple numerical calculations within the same model. As a result, this toy model can be viewed as the near-field-optical counterpart of the usual LASER rate equations.

No abstract

In this paper we present a novel method to increase the spatial resolution of depth images. We combine a deep fully convolutional network with a non-local variational method in a deep primal-dual network. The joint network computes a noise-free, highresolution estimate from a noisy, low-resolution input depth map. Additionally, a highresolution intensity image is used to guide the reconstruction in the network. By unrolling the optimization steps of a first-order primal-dual algorithm and formulating it as a network, we can train our joint method end-to-end. This not only enables us to learn the weights of the fully convolutional network, but also to optimize all parameters of the variational method and its optimization procedure. The training of such a deep network requires a large dataset for supervision. Therefore, we generate high-quality depth maps and corresponding color images with a physically based renderer. In an exhaustive evaluation we show that our method outperforms the state-of-the-art on multiple benchmarks.

No abstract

Thin gold split‐ring resonators are exposed to an aqueous electrolyte. When varying the electrode potential by about 1 V, up to 55‐THz shifts of the resonances centered at around 300 THz are observed. By systematically varying the electrode‐potential window, reversible surface‐charge effects can be separated from irreversible redistributions of gold atoms.