Search

The elementary excitations in metallic glasses (MGs), i.e., β processes that involve hopping between nearby sub-basins, underlie many unusual properties of the amorphous alloys. A high-efficacy prediction of the propensity for those activated processes from solely the atomic positions, however, has remained a daunting challenge. Recently, employing well-designed site environment descriptors and machine learning (ML), notable progress has been made in predicting the propensity for stress-activated β processes (i.e., shear transformations) from the static structure. However, the complex tensorial stress field and direction-dependent activation could induce non-trivial noises in the data, limiting the accuracy of the structure-property mapping learned. Here, we focus on the thermally activated elementary excitations and generate high-quality data in several Cu-Zr MGs, allowing quantitative mapping of the potential energy landscape. After fingerprinting the atomic environment with short- and medium-range interstice distribution, ML can identify the atoms with strong resistance or high compliance to thermal activation, at a high accuracy over ML models for stress-driven activation events. Interestingly, a quantitative “between-task” transferring test reveals that our learnt model can also generalize to predict the propensity of shear transformation. Our dataset is potentially useful for benchmarking future ML models on structure-property relationships in MGs.

Perfluorinate glass ceramics with ultra-low phonon energy are very important optical and photonic materials. Unfortunately, there is no suitable method to obtain transparent perfluorinate glass ceramics due to poor thermal stability of fluoride glass. As a result, wide applications of glass ceramics in advanced infrared systems are restricted. Here, an effective method based on phase-separation engineering is used to develop transparent perfluorinate glass ceramics. In this article, a novel transparent Er3+-doped ZnZrF6-Ba6Zn7F26 perfluorinate glass ceramic was designed and fabricated by phase-separation engineering. The sample exhibits low phonon energy (564 cm−1), ultra-wide transmission range (0.33–8.2 μm, T ≥ 50 %), and strong infrared emission, which is better than that of ZBLAN glass, oxide-, and oxyfluoride-glass ceramics. These good properties of the perfluorinate glass ceramic demonstrate that phase-separation engineering not only offers an effective approach to obtain perfluorinate glass ceramics but also provides wide-ranging opportunities for advanced infrared optical and photonic materials.

Aim This study investigated the impacts of different pre-pregnancy body mass indexes and gestational weight gain on the risk of delivering a high birth weight infant in China. Methods A retrospective cohort study was conducted from 2013 to 2014 in the Maternal and Child Health Care Hospital of Jinan City, Shandong Province and 2415 women who had a singleton birth were included in the study. A logistic regression model and restricted cubic spline regression were used to analyse the association. Findings The risk of delivering a high birth weight infant increases when the mother’s pre-pregnancy body mass index exceeds 24 kg/m2. Compared with women whose pre-pregnancy body mass index was 21 kg/m2, the adjusted risk of delivering a high birth weight infant doubled when the mother’s pre-pregnancy body mass index was 29 kg/m2, and nearly tripled when the mother’s pre-pregnancy body mass index was 31 kg/m2. Compared with women who had a gestational weight gain of 12.0 kg, women having a gestational weight gain of 20.0 kg, 22.0 kg, and 26.0 kg had a 1.7-, 2.2-, and 3.5-fold increased risk of delivering a high birth weight infant. When the mother experiences a gestational weight gain greater than 27 kg, the risk of delivering a high birth weight infant is at least 4-fold greater than that for a mother who has a gestational weight gain of 12.0 kg. Conclusions Proposed strategies to raise public awareness of the risks to infants posed by high maternal pre-pregnancy body mass index and gestational weight gain are required. All clinical recommendations and measures are for all pregnant women, not just overweight and obese pregnant women.

A liquid-crystal-filled photonic crystal fiber (PCF) is proposed for electro-optical modulation. The E7 liquid crystal is precisely filled in one of the innermost air holes of the PCF and forms an in-fiber optical coupler, and the resonance wavelength can be tuned with a sensitivity of 5.594 nm/ V rms when an external voltage is applied. The device can operate as an electro-optical switch/modulator and exhibits response and recovery times of approximately 47 and 24 ms, respectively from 1414 nm to more than 1700 nm. The proposed structure is expected to have potential applications in electric field sensing and wavelength-tunable electro-optical devices.

Class imbalance learning (CIL) is an important branch of machine learning as, in general, it is difficult for classification models to learn from imbalanced data; meanwhile, skewed data distribution frequently exists in various real-world applications. In this paper, we introduce a novel solution of CIL called Probability Density Machine (PDM). First, in the context of Gaussian Naive Bayes (GNB) predictive model, we analyze the reason why imbalanced data distribution makes the performance of predictive model decline in theory and draw a conclusion regarding the impact of class imbalance that is only associated with the prior probability, but does not relate to the conditional probability of training data. Then, in such context, we show the rationality of several traditional CIL techniques. Furthermore, we indicate the drawback of combining GNB with these traditional CIL techniques. Next, profiting from the idea of K-nearest neighbors probability density estimation (KNN-PDE), we propose the PDM which is an improved GNB-based CIL algorithm. Finally, we conduct experiments on lots of class imbalance data sets, and the proposed PDM algorithm shows the promising results.

The breakthroughs in artificial intelligence (AI) have led to a new race in military throughout the world. Even though technologies in AI, like machine learning or big data, are still at their juvenile stage, the technology experts in United States, Europe Union or other countries, have been setting out to transfer AI technologies into the military power, which is doomed to influence all spheres of warfare. To figure out which domain has AI been studied to enhance military capability, the paper summarized 7 major military applications of AI, based on the projects investigated by the defense department in US and Europe governments, finally analyses challenges associated with military employment of AI techniques and draws a brief conclusion.

For a long time, the coal mine auxiliary shaft track transport system of China, including identification harvester loading type and controlling turnout manually, has the problems of low efficiency and high labor cost, and restricting the transportation automation of the coal mine auxiliary shaft. This paper established a classification data set of mine car loading images, proposed a method for mine car loading images based on convolutional neural network. Classifying the images of mine car loading by convolutional neural network technology solves the problems of difficult manual feature extraction and poor adaptability to environmental changes in traditional machine learning methods. Evaluated the actual performance of two typical convolutional neural networks of ResNet and SqueezeNet in the classification of mine car loading on the NVIDIA JETSON TX2 platform. The classification accuracy on the data set reached 97.6% and 98.4%, and the classification speed reached 21fps and 71fps, field test in Gucheng Coal Mine of Linyi Mining Group verified the feasibility of the method.

With the continuous development of computer technology and network communication technology, the machinery manufacturing industry is developing in the direction of network, integration and intelligence. In order to improve the level of digital design and manufacturing, improve the production efficiency, and meet the individual needs of the manufacturing industry, the Digital Factory Platform is introduced into the manufacturing industry. Starting from the concept of Digital Factory, this paper analyzes the overall framework of Digital Factory Platform, constructs a Digital Factory Platform based on Intelligent Manufacturing, provides a feasibility basis for enterprises to build Digital Factory, and provides a model for practical teaching. Through the implementation of the platform, it can cultivate Intelligent Manufacturing related talents with multi-disciplinary comprehensive knowledge and skills.

The temporal response of conventional ZnO nanowire-based photodetectors suffers from the influence of carrier mobility and high electrical resistance. As a result, these devices can be prohibitively slow for some applications, generally with a response time on the order of 1 second. This study presents a novel ZnO nanowire-based microfiber coupler structure for all-optical photodetection without the demand for photocurrent generation. In this design, two waveguides are directly adsorbed by van der Waals forces or electrostatic forces. Compared with the conventional electrical bridge structure, the optical coupling architecture makes the device both compact and stable. In this configuration, resonance dips form in the transmission spectrum when the phase-matching conditions of the two waveguides are satisfied. The measurements of the resulting wavelength shift, low noise levels, and repetition time confirmed that the proposed optical device could operate as a photodetector. The device exhibits superior performance with a sensitivity of 1.657 nm (W cm-2)-1 and a response time of 0.43 ms. In addition, the detector features a simple fabrication as there is no need for extra modification of ZnO nanowires. The resulting photodetection capabilities could provide a new functionality for novel all-optical applications.

For crystalline-silicon (c-Si) solar cells, the interdigitated back contact (IBC) structure has been long known as an efficient way to approach the theoretical limit of efficiency. However, the complexity of fabricating this kind of devices as well as the high dependence on expensive vacuum systems pose concerns about their commercial potential. Here, we demonstrate a novel c-Si IBC solar cell featuring dopant-free heterocontacts for both polarities, i.e. a solution-proceeded PEDOT:PSS film as hole-transporting layer (HTL) and an evaporated magnesium-oxide film as electron-transporting layer (ETL). Our innovatively buried ETL method provides substantial simplification on the architecture and fabrication of the IBC cells and makes it possible to adapt solution-proceeded HTLs while keeping good passivation in gap regions. The IBC solar cell shows an efficiency of 16.3%, with a promising short-circuit current density (Jsc) up to 38.4 mA/cm2. A thorough simulation concerning the influence of pitch size, surface recombination rate (at ETL and gap regions) was conducted, revealing a readily achievable Jsc of 41 mA/ cm 2 and a PCE beyond 22%. Our findings demonstrated a feasibility of using solution method to fabricate high efficiency dopant-free IBC solar cells.

Background Studies have revealed the association between famine in early life and type 2 diabetes mellitus (T2DM) in adulthood. However, studies on the Great Chinese Famine were not conducted nationwide. Because of regional variations in the severity of this famine, the results of regional studies are limited. This study explored associations between famine and T2DM in adulthood in a nationwide study. Methods The present study was performed on 7262 participants who were born between 1 October 1949 and 1 July 1966 using baseline data collected for the China Health And Retirement Longitudinal Study (CHARLS) in 2011. Participants were divided according to birthdate into cohorts with fetal, late, middle, and early childhood exposure and no exposure to famine. Logistic regression models were used to analyze the association between famine exposure in early life and the risk of T2DM and hyperglycemia in adulthood. Results For females, the risk of hyperglycemia was higher for famine-exposed than not exposed cohort (odds ratios [OR] 1.34 and 95% confidence intervals [CIs]: 1.34 [1.04-1.74], 1.48 [1.15-1.90], 1.38 [1.06-1.79], and 1.57 [1.25-1.98] for fetal, early, middle, and late childhood exposure, respectively), and this association was even stronger in female participants who lived in rural areas before the age of 16 years. In males, the risk of T2DM was lower for the early and late childhood exposure than no exposure cohorts (OR [95% CIs]: 0.65 [0.49-0.86] and 0.74 [0.56-0.98], respectively). [Correction added on 23 July 2018, after first online publication: Parts of the above 'Results' section have been corrected to interchange the citation of the words 'late' and 'early'.] CONCLUSION: Exposure to famine during early life can increase the risk of hyperglycemia in female adults, but may decrease the risk of T2DM in males.

This paper makes the first step towards mining citywide traffic congestion correlation by utilizing traffic related information from social media. Traffic congestion correlation mining, namely studying which road segments close to each other are highly likely to occur congestion simultaneously, is especially important to help many real applications, such as traffic prediction, traffic control, and urban transportation planning. Traditional traffic data collected from various sensors and other equipments are costly to obtain and hard to scale up to cover a entire city. With the rising popularity of social media, it is common for the public transportation systems and governments to share real time traffic information with the public through social media. It provides us great opportunities to study the traffic conditions of a city with the rich and easily available online data. However, it is also very difficult to use social media data to mine the citywide traffic congestion correlation due to the following major challenges: (1) Social media data like tweets in Twitter are usually noisy and hard to process, especially those tweets posted by individuals. (2) There lacks a method to study the citywide traffic congestion correlation. In this paper, instead of crawling all the traffic related tweets of a city, we only focus on utilizing the tweets posted by some particular organizations or governments. Tweets posted by them are more accurate and formal, thus it is much easier for traffic information extraction. We regard the traffic congestion correlation mining task as a spatio-temporal frequent pattern mining problem by considering each tweet reporting the traffic congestion of a particular road segment as a spatio-temporal item. A spatio-temporal frequent pattern mining algorithm TC_Apriori is also proposed to discover the road segment co-occurrence patterns in congestion. We use the tweets reporting the traffic information of Chicago to evaluate the proposed approach, and the results show that the proposed approach can effectively discover the road segment co-occurrence patterns in congestion.

Pitch uncertainty and line edge roughness are among the critical quality attributes of a pitch standard and normally the analyses of these two parameters are separate. The analysis of self-traceable Cr atom lithography nano-gratings shows a positive relevance and sensitivity between LER and evaluated standard deviation of pitch. Therefore, LER can be used as an aided pre-evaluation parameter for the pitch calculation method, such as the gravity center method or the zero-crossing points method. The optimization of the nano-grating evaluation method helps to obtain the accurate pitch value with fewer measurements and provide a comprehensive characterization of pitch standards.

Recent decades have witnessed wave and tidal energy technology receiving considerable attention because of their low carbon emissions during electricity production. However, indirect emissions from their entire life cycle should not be ignored. Therefore, life cycle assessment (LCA) has been widely applied as a useful approach to systematically evaluate the environmental performance of wave and tidal energy technologies. This study reviews recent LCA studies on wave and tidal energy systems for stakeholders to understand current status of methodological practice and associated inherent limitations and reveal future research needs for application of LCA on wave and tidal technologies. The conformance of the selected LCAs to ISO 14040 (2006) and 14044 (2006) are critically analyzed in strict accordance with the ISO stepwise methodologies, namely, goal and scope definition, life cycle inventory (LCI) analysis, as well as life cycle impact assessment (LCIA). Our systematically screening of these studies indicates that few of the selected studies are of strict conformance with ISO 14040 and 14044 standards, which makes the results unreliable and thus further reduces the confidence of interested stakeholders. Further, our review indicates that current LCA practice on wave and tidal energies is lacking consideration of temporal variations, which should be addressed in future research, as it causes inaccuracy and uncertainties.

Coal and gas outbursts have occurred in the Zhaozhuang coal mine (a high gas coal mine) since its construction, and these outbursts pose a great threat to lives and property. Therefore, the combination of methane desorption experiments and molecular simulations was adopted to investigate the effects of moisture on the methane desorption characteristics of the Zhaozhuang coal, and the microscopic mechanism was analyzed in this paper. The aim of this work is to provide a method for predicting the effect of moisture on coal seam methane desorption, and quantitatively evaluating the control effect of hydraulic measures on coal and gas outbursts from a molecular perspective. The experimental data (methane desorption amount) were measured under various adsorption equilibrium pressures (0.3 MPa, 0.4 MPa, and 0.5 MPa) and moisture contents (0%, 5%, and 9.8%). Based on the Zhaozhuang coal molecular model, the molecular simulation process was conducted using Materials studio software, which can better match the experimental conditions by setting various pressure parameters (0.3 MPa, 0.4 MPa, and 0.5 MPa) and calculating various numbers of H2O molecules (0, 8, and 16) on the coal molecule. The results show that methane desorption decreases with the increasing moisture content in the experiment. The number of CH4 molecules adsorbed on the coal molecule decreases with the increasing number of H2O molecules in the simulation. This phenomenon can be explained by the competitive adsorption between CH4 and H2O molecule: the interaction between coal molecule and H2O molecules is stronger. The affinity between CH4 molecules and the coal molecule is reduced, and CH4 molecules are less aggregated around the coal molecule because of the presence of H2O molecules.

Semiconductor-based photodetectors have received wide attention. Traditional photodetectors are based on electrical test methods, which inevitably disturbed by dark current noise. In order to overcome this problem, this paper proposes an all-optical photodetection scheme based on a microfiber/SiC-nanowire directional coupling structure, which directly utilizes the refractive index change of SiC nanowire caused by photogenerated carriers instead of the change of photocurrent. The device is fabricated by transferring a single SiC nanowire to a microfiber with diameter of about 1 μm by microscopic operating system. When the device is irradiated by a 266 nm deep ultraviolet laser, its light detection sensitivity reaches up to 103 pm/(W/cm2).

This study was aimed at applying microdialysis to explore cefquinome pharmacokinetics in thigh and plasma of healthy, neutropenic, and Actinobacillus pleuropneumoniae-infected mice. The relative recoveries (RRs) were tested in vitro by dialysis and retrodialysis and in vivo by retrodialysis. ICR mice were randomly divided into four groups: H-40 mg/kg (healthy mice receiving cefquinome at 40 mg/kg), H-160 mg/kg, N-40 mg/kg (neutropenic mice), and I-40 mg/kg (thigh infected-mice with A. pleuropneumoniae). After cefquinome administration, plasma was collected by retro-orbital puncture and thigh dialysate was collected by using a microdialysis probe with Ringer's solution at a perfusion rate of 1.5 L/min. Plasma and thigh dialysate samples were assessed by HPLC-MS/MS and analysed by a non-compartment model. The mean in vivo recoveries in the thigh were 39.35%, 38.59%, and 37.29% for healthy, neutropenic, and infected mice, respectively. The mean plasma protein binding level was 16.40% and was independent of drug concentrations. For all groups, the mean values of the free AUCinf in plasma were higher than those in murine thigh, while the elimination T1/2β for plasma were lower than those for murine thigh. Cefquinome penetration (AUCthigh/AUCplasma) from the plasma to thigh was 0.70, 0.81, 0.49, and 0.98 for H-40, N-40, I-40, and H-160 mg/kg, respectively. These results indicated that infection significantly affected cefquinome pharmacokinetics in murine thigh. In conclusion, we successfully applied a microdialysis methods to evaluate the pharmacokinetics of cefquinome in murine thigh of healthy, neutropenic, and A. pleuropneumonia-infected mice and the pharmacokinetics of cefquinome was obviously affected by infection in thigh.

QingFeiPaiDu decoction (QFPDD) treatment benefits patients with coronavirus disease 2019 (COVID-19). This study aims to elucidate the mechanisms that underlie the anti-inflammatory effects of QFPDD.

Transportation cyber–physical systems (TCPSs) require simulation-based testing and evaluation due to the prohibitive cost of building realistic test beds. Given the transdisciplinary nature of TCPSs, various simulation models and frameworks have been proposed in civil engineering, computer science, and related fields. Traditionally, researchers in different areas have developed their own set of simulation tools, which provide limited capability for TCPS research. In recent years, we have witnessed a growing interest of combining two or more features of traditional simulators to capture the unique characteristics of TCPSs. In this paper, we describe several mainstream simulation models used in transportation, communication, and human-factor studies in TCPS research. Moreover, we present our unique design and implementation of an integrated traffic–driving–network simulator (ITDNS). Finally, we discuss future enhancements that will promote best simulation practices for TCPS research.

In the working environment of aero engine blades, it is necessary to withstand the influence of factors such as aerodynamic stress, vibration stress, centrifugal stress, thermal stress, high temperature oxidation, thermal corrosion, etc. The blades often exhibit vibration fatigue fracture, foreign object damage, corrosion, material defects, and sports damage. This article summarizes the fatigue fracture problems of aero-engine blades, provides a systematic basic theory for the design, innovation and development of aero-engine blades, and has certain reference value.

Image super-resolution based on learning dictionary has recently attracted enormous interests. The learning-based methods usually train a pair of dictionaries from low-resolution and high-resolution image patches, ignoring the fact that patches have different structures. In this paper, we propose to train a set of novel multi-pairs of dictionaries for different categories of patches which clustered by gaussian mixture model, instead of a global dictionary trained from all patches. The multi-pairs of dictionaries via patch prior guided clustering can express structure information of the image patches well. Extensive experimental results prove it has strong robustness in super resolution. Compared with state-of-the-art SR methods, our method demonstrates more pleasant quality of image edge structures and texture.

A surface plasmon resonance (SPR) sensor based on a side-polished single mode fiber coated with polyvinyl alcohol (PVA) is demonstrated for relative humidity (RH) sensing. The SPR sensor exhibits a resonant dip in the transmission spectrum in ambient air after PVA film coating, and the resonant wavelength shifts to longer wavelengths as the thickness of the PVA film increases. When RH changes, the resonant dip of the sensor with different film-thicknesses exhibits interesting characteristics for optical spectrum evolution. For sensors with initial wavelengths between 550 nm and 750 nm, the resonant dip shifts to longer wavelengths with increasing RH. The averaged sensitivity increases firstly and then drops, and shows a maximal sensitivity of 1.01 nm/RH%. Once the initial wavelength of the SPR sensor exceeds 850 nm, an inflection point of the resonant wavelength shift can be observed with RH increasing, and the resonant dip shifts to shorter wavelengths for RH values exceeding this point, and sensitivity as high as &minus, 4.97 nm/RH% can be obtained in the experiment. The sensor is expected to have potential applications in highly sensitive and cost effective humidity sensing.

To explore the in vivo antimicrobial activity of cefquinome against Pasteurella multocida in piglets, a piglet tissue cage infection model was used in this study. After the population of P. multocida reached 107 CFU/mL in a tissue cage, piglets received an intramuscular administration of cefquinome at 0.2, 0.4, 0.8, 1, 2, and 4 mg/kg once daily for 3 days. To assess the tissue cage pharmacokinetics (PKTCF) of cefquinome, tissue cage fluid was collected for cefquinome analysis at 1, 3, 6, 9, 12, and 24 hr after each of the 3 daily drug administrations. Bacteria were counted every 24 hr after drug administration and at 48 and 72 hr after the last administration. Evaluation of the relationship between pharmacokinetic/pharmacodynamic (PK/PD) parameters and the antibacterial effect showed that the surrogate of %T > minimum inhibitory concentration (MIC) (R2 = 0.981) was the best PK/PD index that correlated with effectiveness of cefquinome against P. multocida. The respective values of %T > MIC required for continuous 1/3-log, 1/2-log, and 1-log reductions were 14.23, 34.45, and 73.44%, respectively, during each 24-hr treatment period. In conclusion, cefquinome exhibited a potent antibacterial effect against P. multocida. When %T > MIC reached 73.44%, cefquinome exhibited a bactericidal effect against P. multocida after three successive daily administrations.

As a classic issue in military operation research, weapon target assignment (WTA) has been of interest to researchers for a long time. The purpose of WTA is to determine the best assignment scheme to gain the largest benefit while satisfying a number of constraints deriving from the capability of available platforms and munitions (or weapons), target characteristics, timings for operation and so on. In order to overcome the drawbacks of easily falling into premature convergence and local optimum of existing heuristic algorithms when they are employed to optimize WTA problems, this paper proposes a hybrid genetic algorithm (HGA) which combines an adaptive genetic algorithm (AGA) with an adaptive variable neighborhood search algorithm (AVNSA) to balance the exploration and exploitation ability. In the framework of HGA, AGA is used for wide scope search in the solution space to avoid premature convergence while AVNSA for local search to jump out of the local optimal space. The simulation result demonstrates the effectiveness and feasibility of the proposed HGA in small-scale WTA optimization problems and through a comparison analysis, advantages of the algorithm over the adaptive genetic algorithm, the immune genetic algorithm and the standard genetic algorithm in large-scale WTA optimization problems are revealed.

Convolutional Neural Networks (CNNs) have attracted significant attention in visual recognition. To fully exploit the potential advantages of the CNN models for image classification, this paper introduces several new ideas, ranging from feature generation to classifier selection. We start by transforming the existing CNN models into fully convolutional networks (FCNs). This eliminates the restriction of image resolution and improves the efficiency for data augmentation. Next, we propose a cross pooling strategy to aggregate the top-layer activations, which combines the advantages of average pooling and max pooling, without enlarging the feature dimension. Moreover, we utilize the regularized logistic regression classifier for image classification, and demonstrate that it cooperates better with our feature than the commonly used linear SVM, both in terms of the accuracy and efficiency. By conducting extensive experiments on four frequently benchmarked datasets, we find our proposed scheme usually has better performance relative to other leading approaches which utilize similar-size features, while delivering comparable results with the state-of-the-art methods which depend on much larger dimensional features.

In the present study, there are a number of recognition methods with high recognition accuracy, which are based on deep learning. However, these methods usually have a good effect in a restricted environment, but in the natural environment, the accuracy of face recognition has decreased significantly, especially in the case of occlusion, face recognition will appear inaccurate or unrecognized situation. Based on this, this paper presents a face recognition method based on the deep learning in the natural environment, hoping to achieve robust performance in the natural environment, especially in the case of occlusion. The main contribution of this paper is improving the method of multi-patches by using 4 areas' patches in the face. And in order to have a higher performance, we use a Joint Bayesian (JB) measure in face-verification. Finally, we trained the model by the set of CASIA-WebFace and test it in the Labeled Faces in the Wild (LFW).

Tracking people and objects is a fundamental stage toward many video surveillance systems, for which various trackers have been specifically designed in the past decade. However, it comes to a consensus that there is not any specific tracker that works sufficiently well under all circumstances. Therefore, one potential solution is to deploy multiple trackers, with a tracker output fusion step to boost the overall performance. Subsequently, an intelligent fusion design, yet general and orthogonal to any specific tracker, plays a key role in successful tracking. In this paper, we propose a symbiotic tracker ensemble toward a unified tracking framework, which is based on only the output of each individual tracker, without knowing its specific mechanism. In our approach, all trackers run in parallel, without requiring any details for tracker running, which means that all trackers are treated as black boxes. The proposed symbiotic tracker ensemble framework aims at learning an optimal combination of these tracking results. Our method captures the relation among individual trackers robustly from two aspects. First, the consistency between two successive frames is calculated for each tracker. Then, the pair-wise correlation among different trackers is estimated in the new coming frame by a graph-propagation process. Experimental results on the Caremedia dataset and the Caviar dataset demonstrate the effectiveness of the proposed method, with comparisons to several state-of-the-art methods.

Step-index fluorozirconate (ZBLAN) and fluoroindate (InF3) fibers with all-normal dispersion (ANDi) are proposed for mid-infrared (MIR) nonlinear application. The fibers have a flat and near-zero ANDi profile at the MIR region to generate a highly coherent MIR supercontinuum (SC). A multiwatt coherent MIR SC is allowed to guide in the ANDi ZBLAN and InF3 fibers of 5.5 µm and 6.5 µm core diameters. The MIR SC is simulated in the ANDi fluoride fibers pumped with high peak power at 1960 nm and 2700 nm. The simulation results show that a highly coherent MIR SC spanning from 1100 to 3000 nm with flatness

This Letter highlights a cost-effective, simple, and rapid one-step process leading to the (Sr0.84Lu0.16) F2.16 glass ceramic in a completely new perfluoride system. The mechanism was demonstrated clearly. This material shows high transparency in the UV (0.35 μm) range up to far-IR (10.8 μm). In addition, low phonon energy, as well as good mechanical properties, chemical durability, spectral performance, and long lifetime (7.2 ms) of Er3+:2.7 μm are also possessed by this material. This Letter effectively breaks through the performance limitation of a glass matrix on fluoride crystallites in glass ceramics for the first time, to the best of our knowledge. Meanwhile, it also provides a promising optical material for windows and lasers by a simple and cheap method.

In recent years, extensive research efforts have been dedicated to automatic news content analysis. In this paper, we propose a novel algorithm for anchorperson detection in news video sequences. In this method, the raw news videos are firstly split into shots by a four-threshold method, and the key frames are extracted from each shot. After that, the anchorperson detection is conducted from these key frames by using a clustering-based method based on a statistical distance of Pearson's correlation coefficient. To evaluate the effectiveness of the proposed method, we have conducted experiments on 10 news sequences. In these experiments, the proposed scheme achieves a recall of 0.96 and a precision of 0.97 for anchorperson detection.

As video data generated by users boom continuously, making sense of large scale data archives is considered as a critical challenge for data management. Most existing learning techniques that extract signal-level contents from video data struggle to scale due to efficiency limits. With the development of pervasive positioning techniques, discovering hot topics from multimedia data by their geographical tags has become practical: videos taken by advanced cameras are associated with GPS locations, and geo-tagged videos from YouTube can be identified by their associated GPS locations on Google Maps. It enables us to know the cultures, scenes, and human behaviors from videos based on their spatio-temporal distributions. However, meaningful topic discovery requires an efficient clustering approach, through which coherent topics can be detected according to particular geographical regions without out-of-focus effects. To handle this problem, this paper presents a filter-refinement framework to discover hot topics corresponding to geographical dense regions, and then introduces two novel metrics to refine unbounded hot regions, together with a heuristic method for setting rational thresholds on these metrics. The results of extensive experiments prove that hot topics can be efficiently discovered by our framework, and more compact topics can be achieved after using the novel metrics.

Automatic understanding of human activities is a huge challenge in multimedia analysis field. This challenge is especially critical in small-scale activities, such as finger motions, and activities in complex scenes. For typical camera views, both global feature and local feature analysis methods are unsuitable. To solve this problem, many studies focus on using spatio-temporal features and feature selection methods to get video representation. However, these spatio-temporal features are problematic for two reasons. First, we are not sure whether these features are meaningful foreground or noise. Second, we are unable to foresee where an activity will occur based on these features. Therefore, a biological feature selection method is needed to reorganize these spatio-temporal features and represent the video in a feature space. In this paper, we propose a graph based Co-Attention model to select more efficient features for activity analysis. Without reducing the dimensionality, our Co-Attention model considers the number of interest points. Our model is derived from correlations among individual tiny activities, whose salient regions are identified by combining an integrated top-down and bottom-up visual attention model, and a motion attention model built by spatio-temporal features instead of optical flow directly. Different from typical attention models, the Co-Attention model allows multiple regions of interest in video co-existing for further analysis. Experimental results on the KTH dataset, YouTube dataset and a new tiny activity dataset, Pump dataset which consist of visual observation data from patients operating an infusion pump, validate our activity analysis approach is more effective than state-of-the-art methods.

PEDOT PSS/silicon heterojunction solar cell has recently attracted much attention due to the fact that it can be simply and cost-effectively fabricated. It is crucial to suppress the interfacial recombination rate between silicon (Si) and organic film for improving device efficiency. In this study, we demonstrated a thickness-dependent passivation effect, i.e. the passivation quality over Si substrate was promoted dramatically with increasing the thickness of PEDOT:PSS layer. The effective minority carrier lifetime increased from 32 μs for 50 nm to 360 μs for 200 nm, which corresponds to a change in implied open circuit voltage (V oc-implied) from 545 to 635 mV. Back-junction hybrid solar cells featuring PEDOT:PSS films at rear side were designed to enable adoption of thick PEDOT:PSS layers without having to worry about parasitic absorption, showing a power conversion efficiency (PCE) of 16.3%. Combined with a proper pre-condition on the Si substrate, the back-junction hybrid solar cell with 200 nm PEDOT:PSS layer received an enhanced PCE of 16.8%. In addition, the improved long-term stability for the back-junction device was also observed. The PCE remained 90% (unsealed) after being stored in ambient atmosphere for 30 days and over 80% (sealed) after 150 days.

Background: The multilayer gratings are considered as the potential length-standard-traceable lateral scales for calibrating the next-generation critical dimension scanning electron microscope (CD-SEM) magnification. As a key step in the fabrication of multilayer gratings, selective wet etching determines the final grating structure formation. However, the effects of the etching process parameters on the multilayer gratings in several nanometer scales have not been reported in detail. Aim: By optimizing the process of selective wet etching, we should fabricate high-aspect-ratio and uniform multilayer gratings to obtain high-contrast secondary electron signals and stable secondary electron images while also obtaining measurement accuracy from the small line edge roughness. Approach: Based on the analysis of the important factors in the etching process and SEM and TEM measurement results, we evaluate the effects of ultrasonic agitation, HF acid concentration, etch time, and linewidth scale on the aspect-ratio and uniform of multilayer gratings. Results: We recommend to etching the multilayer films with an HF acid concentration of about 2% during the ultrasonic agitation for uniformity. Moreover, selective wet etching reaction is limited by scale when the linewidth is below 20 nm. Despite the fact that the grating structure is fragile and easy to be broken down, for linewidths of about 10 and 5 nm, the aspect ratio of multilayer gratings can reach about 3 and 2, respectively. Conclusions: By focusing on the optimum conditions of ultrasonic agitation, HF acid concentration, and linewidth scale in the selective wet etching, selective wet etching can be used to fabricate high-aspect-ratio and uniform multilayer gratings with linewidth below 20 nm.

We propose and experimentally demonstrate a highly sensitive gas pressure sensor based on a near-balanced Mach-Zehnder interferometer (MZI) and constructed by hollow-core photonic bandgap fiber (HC-PBF) in this paper. The MZI is simply constructed by fusion splicing two HC-PBFs, which are of slightly different lengths, between two 3-dB couplers. The two output ends of each coupler are approximately equal in length, to ensure that the optical path variations of the MZI only result from the differences in the lengths between the two HC-PBFs. To apply the MZI for gas pressure sensing, a femtosecond laser is employed to drill a micro-channel in one of the two HC-PBF arms. The experiment result shows that the proposed MZI based gas pressure sensor achieves an ultrahigh sensitivity, up to 2.39 nm/kPa, which is two orders of magnitude higher than that of the previously reported MZI-based gas pressure sensors. Additionally, the effects resulting from the absolute length and relative length of the two HC-PBFs on gas pressure sensing performance are also investigated experimentally and theoretically, respectively. The ultra-high sensitivity and ease of fabrication make this device suitable for gas pressure sensing in the field of industrial and environmental safety monitoring.

Line edge roughness (LER) in a one dimensional Si/SiO2 multilayer grating reference material with 20 nm nominal pitch size was investigated. It was shown for the first time that the LER of Si on SiO2 edges was about three times larger than that of SiO2 on Si edges. The asymmetric LERs led to asymmetric uncertainties and it must be considered when using the multilayer grating reference material. Moreover, the origin of asymmetric LER was investigated based on distinguishing contributions of interfacial roughness and interfacial diffusion. The interfacial roughness was determined to be symmetric because the surface roughness of the substrate, the Si single layer, the Si/SiO2 bilayer and the Si/SiO2 multilayer were almost the same with the value about 0.1 nm. Whereas, the interfacial diffusion layer of Si on SiO2 interface was about two times larger than that of SiO2 on Si interface through X-ray reflectivity (XRR) reverse fitting and transmission electron microscopy (TEM) image analysis. The asymmetric interfacial diffusion layer was proposed to be the main reason of the observed asymmetric LERs of the Si/SiO2 multilayer grating reference material.

How to build a good appearance descriptor for tracking target is a basic challenge for long-term robust tracking. In recent research, many tracking methods pay much attention to build one online appearance model and updating by employing special visual features and learning methods. However, one appearance model is not enough to describe the appearance of the target with historical information for long-term tracking task. In this paper, we proposed an online adaptive multiple appearances model to improve the performance. Building appearance model sets, based on Dirichlet Process Mixture Model (DPMM), can make different appearance representations of the tracking target grouped dynamically and in an unsupervised way. Despite the DPMM’s appealing properties, it characterized by computationally intensive inference procedures which often based on Gibbs samplers. However, Gibbs samplers are not suitable in tracking because of high time cost. We proposed an online Bayesian learning algorithm to reliably and efficiently learn a DPMM from scratch through sequential approximation in a streaming fashion to adapt new tracking targets. Experiments on multiple challenging benchmark public dataset demonstrate the proposed tracking algorithm performs 22 % better against the state-of-the-art.

A highly sensitive refractive index sensor based on surface plasmon resonance in a side-polished low-index polymer optical fiber is proposed for biosensing. Benefitting from the low refractive index of the fiber core, the sensitivity of the device can reach ~44567 nm/RIU theoretically for aqueous solutions, at the expense of a lowered upper detection limit that is down to ~1.340. The sensor is fabricated by coating 55-nm-thick Au-film on the polished surface of a graded-index perfluorinated polymer optical fiber. Results show that the sensor exhibits a sensitivity of ~22779 nm/RIU at 1.335 with a figure of merit of 61.2. When employed for glucose sensing, the sensor presents an averaged sensitivity of 24.50 nm/wt%, or 0.46 nm/mM. This device is expected to have potential applications in cost-effective bio- and chemical-sensing.

Tracking an object in long term is still a great challenge in computer vision. Appearance modeling is one of keys to build a good tracker. Much research attention focuses on building an appearance model by employing special features and learning method, especially online learning. However, one model is not enough to describe all historical appearances of the tracking target during a long term tracking task because of view port exchanging, illuminance varying, camera switching, etc. We propose the Adaptive Multiple Appearance Model (AMAM) framework to maintain not one model but appearance model set to solve this problem. Different appearance representations of the tracking target could be employed and grouped unsupervised and modeled by Dirichlet Process Mixture Model (DPMM) automatically. And tracking result can be selected from candidate targets predicted by trackers based on those appearance models by voting and confidence map. Experimental results on multiple public datasets demonstrate the better performance compared with state-of-the-art methods.

A proof of concept performance monitoring system was designed to allow for the realtime assessment of the physiological exertion of Special Forces soldiers during combat. The system was designed to be integrated into the Tactical Assault Light Operator Suit being developed the United States Special Operations Command. It consisted of a suite of sensors integrated into a headset that measures heart rate, respiratory rate, and end tidal CO 2 . The data was presented as a color-coded scale with different colors representing increasing levels of exertion. The prototype was then tested to ensure that it met its design criteria.

This paper addresses temporal synchronization of human actions under multiple view situation. Many researchers focused on frame by frame alignment for sync these multi-view videos, and expolited features such as interesting point trajectory or 3d human motion feature for event detecting individual. However, since background are complex and dynamic in real world, traditional image-based features are not fit for video representation. We explore the approach by using robust spatio-temporal features and self-similarity matrices to represent actions across views. Multiple sequences can be aligned their temporal patch(Sliding window) using the Dynamic Time Warping algorithm hierarchically and measured by meta-action classifiers. Two datasets including the Pump and the Olympic dataset are used as test cases. The methods are showed the effectiveness in experiment and suited general video event dataset.

We demonstrate an egocentric human activity assistant system that has been developed to aid people in doing explicitly encoded motion behavior, such as operating a home infusion pump in sequence. This system is based on a robust multi-camera egocentric human behavior detection approach. This approach detects individual actions in interesting hot regions by spatio-temporal mid-level features, which are built by spatial bag-of-words method in time sliding window. Using a specific infusion pump as a test case, our goal is to detect individual human actions in the operations of a home medical device to see whether the patient is correctly performing the required actions.

Many trackers have been proposed for tracking objects individually in previous research. However, it is still difficult to trust any single tracker over a variety of circumstances. Therefore, it is important to estimate how well each tracker performs and fusion the tracking results. In this paper, we propose a symbiotic black-box tracker (SBB) that learns only from the output of individual trackers, which run in parallel, without any detailed information about these trackers and selects the best one to generate the tracking result. All trackers are considered as black-boxes and SBB learns the best combination scheme for all existing tracking results. SBB estimates confidence scores of these trackers. The confidence score is estimated based on the tracking performance of each tracker and the consistency performance among different trackers. SBB is employed to select the best tracker with the maximum confidence score. Experiments and comparisons conducted on the "Caremedia" dataset and the "Caviar" dataset demonstrate the effectiveness of the proposed method.

In order to reflect the dynamic feature of situations and of peers' behavior more accurately, context information, which contains different attributes of situations, should be considered explicitly when evaluating trust. A context-aware trust establishment framework based on Bayesian networks is proposed. In our proposal, the contextual factors are inferred statistically, not assigned in ad hoc way as most existing solutions. We also propose a Bayesian trust mapping approach between related contexts to help application agents infer the trustworthiness in an unfamiliar context from the trust information in other contexts. We argue that the relevance of contexts is specific to agents and it should not be estimated independent of agents' trust or even fixed as in other frameworks. Our approach makes the relevance measurement implicit together with trust mapping through Bayesian networks. Simulation experiments show that our system can make more accurate trust inferences than ad hoc context-aware systems.

Video shot classification is an important component of content-based video retrieval. It is also a basic step towards video abstract, event detection and content filtering. According to the characters of shots for soccer videos, an integrating color and edge distribution shot classification method is presented. First, A GMM model is trained for grass pixel. And then the grass distribution is computed based on the trained GMM. Due to the sensitivity to light, field and time for color features, edge distribution is computed through canny operator. These features are reasonable complementarities to color features and have no sensitivity to light, field and time. Integrating two types of features embodies main characters of different shot types. One-against-others SVMs are designed for shots classification. Experiments show that our method performs better results.