Single emitters have been considered as sources of single photons in various contexts such as cryptography, quantum computation, spectroscopy, and metrology 1,2,3 . The success of these applications will crucially rely on the efficient directional emission of photons into well-defined modes. To accomplish a high efficiency, researchers have investigated microcavities at cryogenic temperatures 4 , photonic nanowires 5, and near-field coupling to metallic nano-antennas 6 . However, despite an impressive progress, the existing realizations substantially fall short of unity collection efficiency. Here we report on a theoretical and experimental study of a dielectric planar antenna, which uses a layered structure for tailoring the angular emission of a single oriented molecule. We demonstrate a collection efficiency of 96% using a microscope objective at room temperature and obtain record detection rates of about 50 MHz. Our scheme is wavelength-insensitive and can be readily extended to other solid-state emitters such as color centers 7 and semiconductor quantum dots 8 .One of the most powerful and versatile approaches to the generation of single photons exploits the property that a single quantum mechanical two-level system cannot emit two photons simultaneously since each excitation and emission cycle requires a finite time. Unfortunately, such single-photon sources (SPS) are intrinsically inefficient because their radiation spreads over a 4π solid angle and cannot be fully captured by conventional optics. Several years ago, a simple avenue for efficient photon collection was proposed by Koyama et al. in the context of fluorescence microscopy 9 , where emitters were placed at the interface between two media with large refractive index contrast 9,10 . Such a structure can be viewed as a dielectric antenna 11 in which the dipolar radiation of the emitter is funneled into the high-index substrate. The black trace in Fig. 1a shows the angular emission of a dipole sitting close to an interface and oriented perpendicular to it. Despite the strongly modified radiation pattern, one finds that 14% of the light is still lost to the upper half-space, and more importantly, a considerable amount of light is directed to very large angles in the lower substrate, which are not accessible by the collection optics. In this Letter, we remedy these issues by embedding the emitter in a dielectric layer that we engineer on top of the highindex substrate and obtain unprecedented photon collection efficiencies, directionality, and count rates.To provide an intuitive explanation of our antenna design, let us decompose the radiation of a dipolar emitter into plane waves and consider the propagation of each component 12. This

Deep learning is currently an extremely active research area in machine learning and pattern recognition society. It has gained huge successes in a broad area of applications such as speech recognition, computer vision, and natural language processing. With the sheer size of data available today, big data brings big opportunities and transformative potential for various sectors; on the other hand, it also presents unprecedented challenges to harnessing data and information. As the data keeps getting bigger, deep learning is coming to play a key role in providing big data predictive analytics solutions. In this paper, we provide a brief overview of deep learning, and highlight current research efforts and the challenges to big data, as well as the future trends.INDEX TERMS Classifier design and evaluation, feature representation, machine learning, neural nets models, parallel processing.

We report on the assembly of low-loss silica nanowires into functional microphotonics devices on a low-index nondissipative silica aerogel substrate. Using this all-silica technique, we fabricated linear waveguides, waveguide bends, and branch couplers. The devices are significantly smaller than existing comparable devices and have low optical loss, indicating that the all-silica technique presented here has great potential for future applications in optical communication, optical sensing, and high-density optical integration.

No abstract

We devise new optical antennas that reduce the excited-state radiative lifetimes of emitters to the order of 100 femtoseconds while maintaining quantum efficiencies of about 80% at a broadband operation.Here, we combine metallic nanoparticles with planar dielectric structures and exploit design strategies from plasmonic nanoantennas and concepts from Cavity Quantum Electrodynamics to maximize the local density of states and minimize the nonradiative losses incurred by the metallic constituents. The proposed metallo-dielectric hybrid antennas promise important impact on various fundamental and applied research fields, including photophysics, ultrafast plasmonics, bright single photon sources and Raman spectroscopy.