Search

Superconducting nanowire single-photon detector (SNSPD) with near-unity system efficiency is a key enabling, but still elusive technology for numerous quantum fundamental theory verifications and quantum information applications. The key challenge is to have both a near-unity photon-response probability and absorption efficiency simultaneously for the meandered nanowire with a finite filling ratio, which is more crucial for NbN than other superconducting materials (e.g., WSi) with lower transition temperatures. Here, we overcome the above challenge and produce NbN SNSPDs with a record system efficiency by replacing a single-layer nanowire with twin-layer nanowires on a dielectric mirror. The detector at 0.8 K shows a maximal system detection efficiency (SDE) of 98% at 1590 nm and a system efficiency of over 95% in the wavelength range of 1530-1630 nm. Moreover, the detector at 2.1K demonstrates a maximal SDE of 95% at 1550 nm using a compacted two-stage cryocooler. This type of detector also shows the robustness against various parameters, such as the geometrical size of the nanowire, and the spectral bandwidth, enabling a high yield of 73% (36%) with an SDE of >80% (90%) at 2.1K for 45 detectors fabricated in the same run. These SNSPDs made of twin-layer nanowires are of important practical significance for batch production., Comment: 17 pages, 13 figures

Let $\mathfrak{a},\mathfrak{b}$ be two ideals of a commutative noetherian ring $R$ and $M$ a finitely generated $R$-module.~We continue to study $\textrm{f}\textrm{-}\mathrm{grad}_R(\mathfrak{a},\mathfrak{b},M)$ which was introduced in [Bull. Malays. Math. Sci. Soc. 38 (2015) 467--482], some computations and bounds of $\textrm{f}\textrm{-}\mathrm{grad}_R(\mathfrak{a},\mathfrak{b},M)$ are provided.~We also give the structure of $(\mathfrak{a},\mathfrak{b})$-$\mathrm{f}$-modules,~various properties which are analogous to those of Cohen Macaulay modules are discovered., Comment: 9 pages. Comments are welcome

The goal of the article is to get a satisfactory theory of cosupport in the derived category $\mathrm{D}(R)$, this is done by introducing another versions of the "big" and "small" cosupport for complexes. We provide some properties for cosupport that are similar--or rather dual--to those of support for complexes. By examples we show that these versions are differ from the cosupport in [J. Reine Angew. Math. 673 (2012) 161--207]. We also study some relations between the "big" and "small" cosupport and give some computations and comparisons of the "small" support and "small" cosupport., Comment: 14 pages. Comments are welcome

We introduce the notions of Koszul $N$-complex, $\check{\mathrm{C}}$ech $N$-complex and telescope $N$-complex, explicit derived torsion and derived completion functors in the derived category $\mathbf{D}_N(R)$ of $N$-complexes using the $\check{\mathrm{C}}$ech $N$-complex and the telescope $N$-complex. Moreover, we give an equivalence between the category of cohomologically $\mathfrak{a}$-torsion $N$-complexes and the category of cohomologically $\mathfrak{a}$-adic complete $N$-complexes, and prove that over a commutative noetherian ring, via Koszul cohomology, via RHom cohomology (resp. $\otimes$ cohomology) and via local cohomology (resp. derived completion), all yield the same invariant., Comment: comments welcome

The goal of the article is to better understand cosupport in triangulated categories since it is still quite mysterious. We study boundedness of local cohomology and local homology functors using Koszul objects, give some characterizations of cosupport and get some results that, in special cases, recover and generalize the known results about the usual cosupport. Also we include some computations of cosupport, settle the comparison of support and cosupport of cohomologically finite objects. Finally, we assign to any object of the category a subset of $\mathrm{Spec}R$, called the big cosupport.

Segmentation is a prerequisite yet challenging task for medical image analysis. In this paper, we introduce a novel deeply supervised active learning approach for finger bones segmentation. The proposed architecture is fine-tuned in an iterative and incremental learning manner. In each step, the deep supervision mechanism guides the learning process of hidden layers and selects samples to be labeled. Extensive experiments demonstrated that our method achieves competitive segmentation results using less labeled samples as compared with full annotation., Comment: Accepted version to be published in the 42nd IEEE Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2020, Montreal, Canada

Data-driven models are becoming essential parts in modern mechanical systems, commonly used to capture the behavior of various equipment and varying environmental characteristics. Despite the advantages of these data-driven models on excellent adaptivity to high dynamics and aging equipment, they are usually hungry to massive labels over historical data, mostly contributed by human engineers at an extremely high cost. The label demand is now the major limiting factor to modeling accuracy, hindering the fulfillment of visions for applications. Fortunately, domain adaptation enhances the model generalization by utilizing the labelled source data as well as the unlabelled target data and then we can reuse the model on different domains. However, the mainstream domain adaptation methods cannot achieve ideal performance on time series data, because most of them focus on static samples and even the existing time series domain adaptation methods ignore the properties of time series data, such as temporal causal mechanism. In this paper, we assume that causal mechanism is invariant and present our Causal Mechanism Transfer Network(CMTN) for time series domain adaptation. By capturing and transferring the dynamic and temporal causal mechanism of multivariate time series data and alleviating the time lags and different value ranges among different machines, CMTN allows the data-driven models to exploit existing data and labels from similar systems, such that the resulting model on a new system is highly reliable even with very limited data. We report our empirical results and lessons learned from two real-world case studies, on chiller plant energy optimization and boiler fault detection, which outperforms the existing state-of-the-art method.

High-precision time transfer is of fundamental interest in physics and metrology. Quantum time transfer technologies that use frequency-entangled pulses and their coincidence detection have been proposed, offering potential enhancements in precision and better guarantees of security. In this paper, we describe a fiber-optic two-way quantum time transfer experiment. Using quantum nonlocal dispersion cancellation, time transfer over a 20-km fiber link achieves a time deviation of 922 fs over 5 s and 45 fs over 40960 s. The time transfer accuracy as a function of fiber lengths from 15 m to 20 km is also investigated, and an uncertainty of 2.46 ps in standard deviation is observed. In comparison with its classical counterparts, the fiber-optic two-way quantum time transfer setup shows appreciable improvement, and further enhancements could be obtained by using new event timers with sub-picosecond precision and single-photon detectors with lower timing jitter for optimized coincidence detection. Combined with its security advantages, the femtosecond-scale two-way quantum time transfer is expected to have numerous applications in high-precision middle-haul synchronization systems., Comment: 10 pages, 7 figures

Refrigeration and chiller optimization is an important and well studied topic in mechanical engineering, mostly taking advantage of physical models, designed on top of over-simplified assumptions, over the equipments. Conventional optimization techniques using physical models make decisions of online parameter tuning, based on very limited information of hardware specifications and external conditions, e.g., outdoor weather. In recent years, new generation of sensors is becoming essential part of new chiller plants, for the first time allowing the system administrators to continuously monitor the running status of all equipments in a timely and accurate way. The explosive growth of data flowing to databases, driven by the increasing analytical power by machine learning and data mining, unveils new possibilities of data-driven approaches for real-time chiller plant optimization. This paper presents our research and industrial experience on the adoption of data models and optimizations on chiller plant and discusses the lessons learnt from our practice on real world plants. Instead of employing complex machine learning models, we emphasize the incorporation of appropriate domain knowledge into data analysis tools, which turns out to be the key performance improver over state-of-the-art deep learning techniques by a significant margin. Our empirical evaluation on a real world chiller plant achieves savings by more than 7% on daily power consumption., Comment: CIKM2017. Proceedings of the 26th ACM International Conference on Information and Knowledge Management. 2017