Search

Modifying the wide band gap semiconductor hexagonal boron nitride (hBN) can bring new chances in photonics. By virtue of the solvothermal/hydrothermal oxidation or functionalization, hBN can be converted into fluorescent nanodots. Until now, it has been a big challenge to drily oxidize hBN and turn it into bright fluorescent structures due to its superior chemical stability. Here, we report the oxidation of multilayer hBN into fluorescent structures by ultraviolet (UV, λ = 172 nm) photodissociated directional oxygen radical [O(

More and more enterprises are migrating services and data into virtual machine (VM) using base and increment, and the files in the VM may contain critical data. Therefore, it is necessary to build a trusted environment to enhance the security of the files in VM by the integrity measurement methods (IMMs). In order to simplify the management of the files in integrity measurement, the security manager needs to put these files into one region and then to measures the integrity of the region. If the region is integrity, the files in that region are integrity too. Currently, the traditional IMMs are all based on the message digest algorithms that use too much amount of data, spend much time, and make the performance of VMs degraded. Aiming to address those problems, we propose a novel IMM based on copy-on-write for the region, called RIMM. The method puts the region into the base image, gets the integrity information of the region from the base image by the structures of image and region, and periodically measures the integrity of the region in the increment image by the structure of image and copy-on-write. It is transparent for VM and can eliminate the semantic gap. The evaluation shows that the RIMM can significantly reduce the amount of data and time in the process of integrity measurement. For example, (1) when the region size is 100MB, the amount of data used by RIMM is about 400 times smaller than that used by IMMs based on MD5; the time spent by RIMM is about 600 times less than that used by IMMs based on MD5. (2) when the region size is 9GB, the amount of data used by RIMM is about 29000 times smaller than that used by IMMs based on MD5; the time spent by RIMM is about 3864 times less than that used by IMMs based on MD5.

Whispering gallery mode (WGM) optical microcavities are important building blocks in photonic integrated circuits. Operation of such cavities on specific lower- or higher- order transverse modes has much interest in application perspectives. Here, we demonstrate active mode selection by introducing defects in lithium niobate on insulator microdisks. A focused ion beam is applied to precisely inscribe nano slits into the perimeter of the microdisk. The transmission spectra can be significantly thinned out without severe quality factor degradation. Either fundamental or high-order transverse WGMs can be retained by properly designing the size and location of the defects. The approach may have promising applications in single-mode lasing and nonlinear optics.

Since the existing works in data stream anomaly detection mainly locate abnormal objects from data sequences instead of detecting the abnormal state of a monitored object in real-time. And they need a set of empirical predefined thresholds, which may affect the flexibility of the detection models. These limitations make the existing solutions no longer suitable for the real-time detection of the changes in the status of the monitored object. In this paper, we propose a new Abnormal conditions Detection approach (ADCMO) on the Continuously Monitored Object. ADCMO incorporates the data distribution property to carry out the anomaly quantification for the data in the data stream, and then the abnormal conditions of the monitored object are quantified by the anomaly information of the current data and historical data. Finally, the anomaly state is adaptively determined by the distribution of anomaly coefficients. ADCMO can detect the changes in the status by calculating the outlier coefficient of the object at each moment and adaptively make the abnormal early warning. The experiments show that ADCMO can do this any-time, dynamically, efficiently and effectively.

The integrity measurement method is used to detect whether the files are tampered with and to build a trusted environment. It can improve the security of virtual machines using base and increment image. Currently, the traditional integrity measurement methods (MDA-IMM) are based on the message digest algorithm with high computational complexity and heavy data. As a result, the MDA-IMM consumes a lot of I/O resources and spends too much time. To address those issues, we propose a novel method (COW-IMM) based on copy-on-write for the files in base image, the precondition is that, there is one-to-one correspondence between the cluster of image and the logic block of file system, and both of them have the same size. The COW-IMM gets the information of files for integrity measurement from base image and measures the integrity of files in increment image. We implement a prototype based on KVM, Qcow2 image, and Ext4. The algorithm analysis shows that, the volume of data used by COW-IMM is 512 times smaller than that used by MDA-IMM at least, if the file size is the same. The experimental evaluations show that, the speed of COW-IMM is faster and faster than that of MDA-IMM with the increment of file size. For example, when the file size is 0.1M, the speed of COW-IMM is about 10 times faster than that of MDA-IMM; when the file size is 90M, the speed of COW-IMM is about 592 times faster than that of MDA-IMM.

Although the genetic algorithm has been widely used in the polarity optimization of mixed polarity Reed-Muller (MPRM) logic circuits, few studies have taken into account the polarity conversion sequence. In order to improve the efficiency of polarity optimization of MPRM logic circuits, we propose an efficient and fast polarity optimization approach (FPOA) considering the polarity conversion sequence. The main idea behind the FPOA is that, firstly, the best polarity conversion sequence of the polarity set waiting for evaluation is obtained by using the proposed hybrid genetic algorithm (HGA); secondly, each of polarity in the polarity set is converted according to the best polarity conversion sequence obtained by HGA. Our proposed FPOA is implemented in C and a comparative analysis has been presented for MCNC benchmark circuits. The experimental results show that for the circuits with more variables, the FPOA is highly effective in improving the efficiency of polarity optimization of MPRM logic circuits compared with the traditional polarity optimization approach which neglects the polarity conversion sequence and the improved polarity optimization approach with heuristic technique.

Graphene-based p-n junction photodiodes have a potential application prospect in photodetection due to their broadband spectral response, large operating bandwidth, and mechanical flexibility. Here, we report an ultraviolet (UV) rewritable p-n junction photodiode in a configuration of graphene coated with an amorphous phosphor of 4-bromo-1,8-naphthalic anhydride derivative polymer (poly-BrNpA). Under moderate UV irradiation, occurrence of photoisomerization reaction in the poly-BrNpA film leads to its drastically modified optical characteristics and a concurrent n-type doping in the underneath graphene. Meanwhile, the poly-BrNpA film, highly sensitive to water molecules, has a capability of restoring graphene to its initial p-type doping status by means of water adsorption. Based on these findings, a lateral graphene/poly-BrNpA p-n junction photodiode, responsive to visible light at the junction interface, can be written by UV irradiation and then erased via water adsorption. The p-n junction photodiode is rewritable upon such repetitive loops showing repeatable optoelectronic properties. This study provides a new scheme and perspective of making graphene-based rewritable p-n junction photodiodes in a flexible and controllable way, and it may contribute to expanding new families of optoelectronic devices based on two-dimensional materials.

This paper presents a fully convolutional neural network to tackle the mapping between single view RGB images and depth maps. To regress the depth maps from monocular images, we leverage deep short skip connections in residual learning for extracting features rather than using hand-crafted features. We further propose long skip connections in up-sampling stage to reuse the feature maps which is proved to enhance the result experimentally. To show the impact of loss functions in monocular depth map predictions, we train our model with kind of loss functions and compare the results qualitatively and quantitatively. The proposed model outperforms all current state-of-the-art results with less training data as well as less than half of training epochs in two standard benchmark data sets without any post-processing procedures or other refinement steps.

Recently, a hybrid graphene/poly-BrNpA phosphor structure was used to fabricate rewritable p-n junction photodiode responsive to visible light. In this process, photoisomerization occurred under moderate ultraviolet (UV) irradiation (λ = 365 nm, optical intensity IUV = 1.318 μW μm−2), and the subsequent water (H2O) molecule adsorption facilitate charge transfer at the interface. In this study, we explore the feasibility of making UV photodetector based on the hybrid graphene/poly-BrNpA field-effect transistor. The incident UV light (λ = 365 nm, IUV≤1.81 × 10−4 μW μm−2), which is at least four orders weaker than that used to induce observable photoisomerization in the phosphor, is mostly absorbed to create photo-induced electron-hole pairs. The subsequent separation and transfer of the photo-induced electrons to graphene contribute to high photocurrent by the photogating effect. The photodetector has a responsivity and specific detectivity up to 159 A W−1 and 1.55 × 1011 Jones, respectively, when biased at VDS = 1 V. The photodetector has a slow response, especially to UV shut-off with the fall time τF up to 881 s. By virtue of H2O molecule adsorption, τF can be reduced to 6.6 s by accelerating the reversion and recombination of electrons with the photo-induced holes trapped in the poly-BrNpA film through introducing defect states. This work brings a new perspective for developing photodetectors through combining the functional two-dimensional materials and phosphors.

This paper introduces HSAStore, a newly designed storage system whose goal goes to build an efficient storage system for computing systems that containing large-scale intermediate data. HSAStore involves three sub-systems to work together, a distributed file system which is in charge of the intermediate data, a centralized Network Attached Storage (NAS) which stores the raw input data and the results data, and a local file system that serves for the local data. The HSAStore takes full advantage of the network bandwidth of the computing cluster. Moreover, HSAStore adopts the distributed file system, so it is helpful for efficient execution of parallel programs. Experiments show that HSAStore has a significant improvement on efficiency in the computing systems that containing large-scale intermediate data.

This paper proposes a fast and high performance classifier model for sentiment analysis of textual reviews. The key contribution is three fold. First, a two-level classifier model consists of three base classifiers is proposed, and theory proves that the model could be better than the strongest classifier among the base classifiers in both classification performance and time cost of predict. Second, this paper proposes a lexicon-based classifier as a base classifier using a new part of speech (POS) which is called “weaken words”. Finally, we implemented several two-level classifiers by combining the lexicon-based classifier with several machine learning classifiers. Experiments on Chinese reviews dataset show that the two-level classifier model is effective and efficient.

Since the Local Outlier Factor (LOF) was first proposed, there is a large family of approaches that is derived from it. For the reason that the existing local outliers detection approaches only focus on the extent of overall separation between an object and its neighbors, and do not pay attention to the degree of dispersion between them, the precision of these approaches will be seriously affected in the scattered data sets for outlier detection. In this paper, we redefine the local outliers by combining the degree of dispersion of the object and its neighbors, and propose a new local outlier detection approach (N2DLOF). Compared to conventional approaches, the outliers obtained by N2DLOF are more sensitive to the degree of anomaly of the scattered data sets. Experiments show that our approach has a significant improvement on outlier detection precision in the case of scattered datasets with similar time complexity. In short, we extend the ecosystem of the local outlier detection approaches from a new perspective.

When assisted with an inhomogenous vertical magnetic field, ultraviolet (UV) ozonation turns directional and is testified to be applicable to graphene patterning. Using a more cost-effective low-pressure mercury lamp, we further explore the underlying working mechanism by changing oxygen content, introducing reactive ozone or inert nitrogen molecules, and study the lateral under-oxidation impeded Dirac point shifts for a graphene field-effect transistor under UV irradiation. The paramagnetic oxygen molecule X3Σ−g accelerates toward graphene with the magnetic moment aligned parallel to the magnetic field. The O(3P) atoms, stemming from such a directional oxygen molecule, have a high initial velocity before being further accelerated, and therefore enhance the oxidation capability compared with those from weak diamagnetic ozone molecules. Intermolecular or atomic-molecular collisions between the high-speed oxygen molecules/atoms and the randomly moved weak diamagnetic molecules, including nitrogen and ozone, appear crucial in deteriorating graphene patterning by increasing the lateral under-oxidation. This study may shed light on our understanding of graphene patterning by magnetic-assisted UV ozonation.

Developing an alternative method for fabricating microscale graphene patterns that overcomes the obstacles of organic contamination, linewidth resolution, and substrate damaging is paramount for applications in optoelectronics. Here we propose to pattern chemical vapor deposition grown graphene film through a stencil mask by magnetic-assisted ultraviolet (UV) ozonation under irradiation of a xenon excimer lamp. In this process, the paramagnetic oxygen molecules and photochemically generated oxygen radicals are magnetized and attracted in an inhomogenous external magnetic field. As a consequence, their random motions convert into directional, which can greatly modify or enhance the quality of graphene patterns. Using a ferromagnetic steel mask, an approximately vertical magnetic-field-assisted UV ozonation (BZ = 0.31 T, ∇BZ = 90 T · m−1) has a capability of patterning graphene microstructures with a line width of 29 μm and lateral under-oxidation less than 4 μm. Our approach is applicable to patterning graphene field-effect transistor arrays, and it can be a promising solution toward resist-free, substrate non-damaging, and cost effective microscale patterning of graphene film.