Search

To address the problems in the existing video super-resolution methods, such as noise, over smooth and visual artifacts, which are caused by reliance on limited external training or mismatch of internal similarity instances, this study proposes a video super-resolution reconstruction algorithm based on deep learning and spatio-temporal feature similarity (DLSS-VSR). The video super-resolution reconstruction mechanism with joint internal and external constraints is established utilizing both external deep correlation mapping learning and internal spatio-temporal nonlocal self-similarity prior constraint. A deep learning model based on deep convolutional neural network is constructed to learn the nonlinear correlation mapping between low-resolution and high-resolution video frame patches. A spatio-temporal feature similarity calculation method is proposed, which considers both internal video spatio-temporal self-similarity and external clean nonlocal similarity. For the internal spatio-temporal feature self-similarity, we improve the accuracy and robustness of similarity matching by proposing a similarity measure strategy based on spatio-temporal moment feature similarity and structural similarity. The external nonlocal similarity prior constraint is learned by patch group-based Gaussian mixture model. The time efficiency for spatio-temporal similarity matching is further improved based on saliency detection and region correlation judgment strategy. Experimental results demonstrate that the DLSS-VSR achieves competitive super-resolution quality compared to other state-of-the-art algorithms.

Phase picking is a critical step in the analysis of microseismic data. However, there is a lack of relevant methods for S-phase picking. In this letter, we pick the P-phase and extract the data around the S-phase arrival firstly based on previous studies. Then, we propose a method called time series segmentation clustering (TSSC) which is based on the K-means algorithm to pick the S-phase. In the TSSC method, we construct a feature vector of the microseismic signal and standardize the feature vector first. Then, we modify the K-means objective function into a function of time. We subsequently obtain the S-phase arrival corresponding to the minimum value of the objective function. We tested the objective function of the TSSC method on synthetic signals and microseismic signals, and the results show that the method is reliable. This automatic picking algorithm has been used to pick 1763 S-phase arrivals from microseismic signals. In comparison with manual picking results, a statistical analysis shows that this method is feasible.

Background: RNA-binding protein plays an important role in regulating splicing, RNA transport, and other post-transcriptional processes, identifying special RNA binding domains, and interacting with RNA. Objective: This paper proposes a deep learning framework, DeepFusion-RBP, composed of three submodels. A sliding window is used to obtain sub-sequences, local features are obtained, and then the model is customized for each feature. Methods: The main advantage of this research is using the sliding window method to cut the original sequence. While expanding the data set, this method avoids filling in too much meaningless data. Then, the model is customized for each feature to accurately perform RNA binding protein classification, with specific methods such as LSTM, Conv1D, Amino acid embedding, etc. Results: To test whether the customized model can improve the final prediction effect, we used different combinations of sub-models and test sets of different lengths. The prediction ACC, F1-score and MCC of DeepFusion-RBP are 92.62%, 91.29%, and 84.96%, respectively, with cross-validation. At the same time, DeepFusion-RBP also showed excellent performance on three independent verification sets. Conclusion: The results of 10-fold cross-validation and the independent verification set tests both suggested that the proposed models for different features and intercepting sub-sequences produce a certain improvement in the prediction effect of the model. The data supporting the findings of the article are available at https://github.com/mmwangxu/DeepFusion-RBP-tool.

With the enhancement of the performance of cloud servers, face recognition applications are becoming more and more popular, but it also has some security problems, such as user privacy data leakage. This article proposes a face recognition scheme based on homomorphic encryption in cloud environment. The article first uses the MTCNN algorithm to detect face and correct the data and extracts the face feature vector through the FaceNet algorithm. Then, the article encrypts the facial features with the CKKS homomorphic encryption scheme and builds a database of the encrypted facial feature in the cloud server. The process of face recognition is as follows: calculate the distance between the encrypted feature vectors and the maximum value of the ciphertext result, decrypt it, and compare the threshold to determine whether it is a person. The experimental results show that when the scheme is based on the LFW data set, the threshold is 1.1236, and the recognition accuracy in the ciphertext is 94.8837%, which proves the reliability of the proposed scheme.

The performance of image super-resolution (SR) have been greatly improved with deep convolution neural network (CNN). Despite image SR targets at recovering high-frequency details, most SR methods still focus on generating high-level features via a deep and wide network. They lack the discriminative ability of high-frequency information hidden in the abundant CNN features, thus hindering CNNs to yield better SR results. To tackle this issue, we propose two new attention mechanism: context weighted channel attention (CWCA) and persistent spatial attention (PSA). They modulate abundant features by suppressing the useless features and enhancing the interested ones in a channel-and-spatial manner. The network is then enabled to concentrate more on informative features closely related to the high-frequency components of an image. Furthermore, we propose enhanced attention residual groups with dense connection (EARG-D) to capture not only short-term information but also long-term information to maintain more useful features. Finally, we construct a deep enhanced context attention super resolution network (EASR) for better image reconstruction. Quantitative and qualitative experiments well demonstrate that our proposed method performs better than existing state-of-the-art SR methods.

In the era of coronavirus disease 2019 (COVID-19) pandemic, imported COVID-19 cases pose great challenges to many countries. Chest CT examination is considered to be complementary to nucleic acid test for COVID-19 detection and diagnosis. We report the first community infected COVID-19 patient by an imported case in Beijing, which manifested as nodular lesions on chest CT imaging at the early stage. Deep Learning (DL)-based diagnostic systems quantitatively monitored the progress of pulmonary lesions in 6 days and timely made alert for suspected pneumonia, so that prompt medical isolation was taken. The patient was confirmed as COVID-19 case after nucleic acid test, for which the community transmission was prevented timely. The roles of DL-assisted diagnosis in helping radiologists screening suspected COVID cases were discussed.

Objective This article aims to pay attention to the latest research on the expression, activation and function of hypoxia-inducible factor-2α (HIF-2α) under hypoxia and non-hypoxia conditions, and summarizes the current knowledge about the interaction between hypoxia-inducible factor-2 and angiogenesis, hoping to understand its actions in physiology and disease, with the goal of providing a new strategy for the diagnosis and treatment of wounds. Background Wound healing is a complex and continuous process, involving coagulation, inflammation, angiogenesis, new tissue formation and extracellular matrix remodeling. Of these, angiogenesis is an essential step. One of the main reasons for non-healing or delayed healing of wounds in peripheral vascular diseases and diabetes is the reduced ability to regenerate microvessels through the process of angiogenesis, which has become the focus of new methods for treating chronic wounds. HIF-2α regulates many aspects of angiogenesis, including vascular maturation, cell migration, proliferation and metastasis. Methods Throughout extensive search of PubMed, summarize the medical research on HIF-2α to 2020. Conclusions HIF-2α is necessary for normal embryonic development by stimulating the expression of angiogenic factors, such as vascular endothelial growth factor (VEGF). It is essential for the formation of new blood vessels in physiological and pathophysiological environments. Targeting HIF-2α in wound healing has much clinical significance for tissue repair.

Background: To date, no study has been performed on the relationship between depression and perimenopausal symptoms using the Path model in Iran. Objective: This study aims to assess the relationship between depression and menopause symptoms using the PATH model. Methods: A cross-sectional study was conducted for 4 months, in 2019, and 332 menopausal women were enrolled in 14 public health centers. Data collection instruments included questionnaires of demographic characteristics, anxiety instrument, Menopause Rating Scale, and the Beck Depression short Inventory (BDI-S). The hypothesized mediators were assessed using Structural Equation Modeling (SEM). The model included depression symptoms (dependent variable), anxiety (mediators), number of alive children (moderator), and menopausal symptoms (independent variable). Analyses were conducted by SPSS 16 and AMOS 20. Results: There was a significant positive correlation between menopause symptoms with total anxiety (Rs=0.52), state anxiety (Rs=0.47), trait anxiety (Rs=0.46), and depression (Rs=0.54). Depression had a significant positive correlation with total anxiety (Rs=0.64), trait anxiety (Rs=0.58), and state anxiety (Rs=0.59). Also, the total mediating effect of total anxiety and depression and the total moderating effect of the number of alive children were significant on menopause symptoms. Conclusion: According to the results of the present study, there is an association between depression and menopausal symptoms. Therefore, health providers should pay more attention to the psychological health of women during the perimenopausal period. The predictor factors should be considered in both intervention and clinical assessment of menopause women.

A simple FBG Fabry-Perot sensor system with high sensitivity based on fiber laser beat frequency and Vernier effect is proposed and demonstrated. The sensor head is a Fabry-Perot (F-P) resonant cavity formed by two identical fiber Bragg gratings (FBGs), and the F-P cavity is a part of the laser cavity as a mirror. Vernier effect can be realized by choosing appropriate F-P cavity length and laser cavity length. So we achieved the sensitivity multiplication of the sensor, with sensitivity of 299.17 kHz/°C. The sensing information can be obtained by tracking the envelope of beat frequency signals generated by Vernier effect. This sensor system has the advantages of low cost, simple structure and high sensitivity.

Drug-induced acute liver injury (DIALI) is increasingly recognized as a significant cause of acute liver injury (ALI), which is characterized by a rapid loss of hepatocyte function in patients without pre-existing liver diseases. Evaluation of corresponding biomarkers, including alanine transaminase and aspartate amino transferase, is available as a diagnostic tool for hepatotoxicity. However, these blood tests have certain limitations: (1) they are generally not available for early estimation; (2) it is difficult to visualize and identify hepatotoxicity unambiguously in real-time; and (3) the biomarkers are not unique and are usually influenced by a variety of diseases, leading to potential false results. It is of grave importance and burgeoning demand to develop an early diagnostic approach for such diseases, but the ideal toolkit remains an unresolved challenge.As an alternative, molecular optical probes (fluorescence, chemiluminescence, bioluminescence, etc.) display a lot of advantages, such as high sensitivity, noninvasive fast analysis, and real-time in situ detection. They have emerged as potent and promising tools for the biomedical study of DIALI in living system. Until now, a number of optical probes for DIALI have been reported with some great potential for clinical trials. However, most of the probes still suffer from false signals because of the limitations in clinical application, including poor selectivity, low sensitivity, and biocompatibility. One key challenge that probes face in the ALI environment is the excessive exposure to reactive oxygen/nitrogen species and diffusivity, which may lead to false-positive or negative signals.Our group has employed multiple rational approaches to engineer high-performance optical probes for DIALI. With such development, we have successfully achieved the accurate detection of DIALI with minimal false signals both ex vivo and in vivo. While marching firmly toward understanding the biogenesis and progression of DIALI, we ultimately aim at the early stage clinical diagnosis of the disease, as well as mechanism understanding for clinical trials. In this Account, we summarize and present our three new approaches for the development of high-fidelity optical probes: (1) a combined screening and rational design strategy, (2) a double-locked probe design strategy, and (3) in situ imaging based on the release of a precipitating fluorochrome strategy. Using these strategies, we have formulated probes for a range of biological species that are biomarkers of DIALI, including reactive nitrogen species (ONOO-), reactive sulfur species (H2S and H2Sn), and enzymes (LAP, MAO, and ALP). We have highlighted the rationale for our design and screening strategy and methods to achieve high-fidelity optical probes. Some recent examples of optical probes developed by our laboratory and collaborations are mainly illustrated herein. We anticipate the strategies summarized here to inspire future molecular optical probe design, to contribute to studies of the detailed molecular mechanisms underlying liver diseases, and to improve the efficiency of the diagnosis and treatment of these diseases in clinical settings.

The seismic performance of the precast shear wall was improved by using a newly developed reinforced tenon to strengthen the precast joint, which is used to bear the shear force of the precast joint and reduce the effect of dowel action on the vertical connecting steel bars. The vertical connecting steel bars were only used to bear tensile and compressive stress and consume seismic energy. The seismic performance of reinforced tenon precast shear wall was investigated by quasi-static tests on one reinforced tenon precast shear wall specimen and one ordinary flat seam precast shear wall specimen. This study investigated the crack distribution and failure modes, precast joint slip of the two specimens, and the seismic performance parameters such as bearing capacity, stiffness degradation, ductility performance and energy dissipation capacity. The test results demonstrated that the reinforced tenon effectively reduced the slip of the joint surface and exhibited good ductility and energy dissipation capacity than that of the ordinary flat seam specimen, while the failure mode, bearing capacity, and stiffness were similar to that of the flat seam specimen.

Understanding customer mobility patterns to commercial districts is crucial for urban planning, facility management, and business strategies. Trade areas are a widely applied measure to quantify where the visitors are from. Traditional trade area analysis is limited to small-scale or store-level studies, because information such as visits to competitor commercial entities and place of residence is collected by labour-intensive questionnaires or heavily biased location-based social media data. In this article, we propose CellTradeMap, a novel district-level trade area analysis framework using mobile flow records (MFRs), a type of fine-grained cellular network data. We show that compared to traditional cellular data and social network check-in data, MFRs can model customer mobility patterns comprehensively at urban scale. CellTradeMap extracts robust location information from the irregularly sampled, noisy MFRs, adapts the generic trade area analysis framework to incorporate cellular data, and enhances the original trade area model with cellular-based features. We evaluate CellTradeMap on two large-scale cellular network datasets covering 3.5 million and 1.8 million mobile phone users in two metropolis in China, respectively. Experimental results show that the trade areas extracted by CellTradeMap are aligned with domain knowledge and CellTradeMap can model trade areas with a high predictive accuracy.

When a transformer is running, the insulation oil gradually becomes degraded due to various factors such as electricity and heat, during which low molecular weight gases are formed upon decomposition. By analyzing the gas content in the oil, the state of the insulation oil can be identified, thereby providing an effective basis for transformer fault analysis and diagnosis. However, when the types and severity of the internal faults inside the transformer are different, the components and contents of the gases dissolved in the oil also differ. Sometimes when the types of faults are the same, the gas content may be different. Therefore, if the characteristic of the specific gas content corresponding to a type of fault can be obtained, the transformer faults can be accurately identified according to the contents of different gases, which can most efficiently ensure the reliable operation and maintenance of power transformers. In this paper, an identification method based on the deep transfer network was proposed. According to this method, by studying the existing gas contents and states based on a large number of training data to discover their characteristics, the states of gases in the oil can be precisely identified. Experiments have proved the effectiveness of this method. The ability of state identification by this method is far superior to that of the other existing methods.

Background Chronic obstructive pulmonary disease (COPD) is a chronic airway inflammatory disease characterized by irreversible airflow obstruction. Pathogenic mechanisms underlying COPD remain largely unknown. Objective The current study was designed to explore serum concentration of hypoxia-inducible factor 1α (HIF-1α) in stable COPD patients and the potential effect of Lycium barbarum polysaccharides (LBP) on HIF-1α protein expression. Methods Serum HIF-1α was quantified by ELISA in 102 stable COPD patients before and after 2-week orally taken LBP (100 mL/time, twice daily, 5-15 mg/mL). Correlation of serum LBP and lung function (FEV1%) or blood gas (PO2 and PCO2) was also analyzed. As a control, 105 healthy subjects were also enrolled into this study. Results Serum concentration of HIF-1α was significantly higher in the stable COPD patients (37.34 ± 7.20 pg/mL) than that in the healthy subjects (29.55 ± 9.66 pg/mL, P

Urea is an important molecule in physiology and industry. The detection of urea by electrochemical sensors has broad application prospects in food detection and kidney disease screening. This paper introduces the preparation method of MWCNT/PANi electronic ink and its composite electrode, which was applied to the detection of urea. The linear range of the MWCNT/PANi composite electrode was increased by adjusting the concentration of MWCNT. The maximum linear range was 10 to 100 μM when 5.6 mg MWCNT was added. The sensitivity of the electrode was 2.435 $$\times$$ 10−4 mA mM−1 cm−2 and the detection limit was 53.9 nM. At the same time, an algorithm was proposed to identify and analyze the data collected by the electrochemical workstation in view of the errors existing in the calculation of the electrochemical sensor when analyzing the optimal response voltage of the electrode. This method was used to set constraints and screen the optimal working voltage. Experimental results show that this method has certain advantages.

The nondatum complex component is a complex component and does not have reliable positioning/measurement datum during the modification process. With the rapid development of the aerospace industry, the manual manner of shape trimming for the nondatum complex component needs to be replaced by laser automatic shape trimming. This study develops a vision-based, online, and high-precision method for measuring the fixtured pose error of the component. Limited by burrs, deformation, and complex structure, the feature-based method is not applicable to this component. The present method is based on iterative optimization and not required feature extraction. To improve pose measurement efficiency and accuracy, the optimization of the reference model can effectively reduce the error caused by deformation. The experiment results showed that the proposed method is fast and accurate, and satisfied the measurement requirement of the fixtured component’s pose in six degrees-of-freedom.

To develop a high-accuracy method for predicting SCB based on the analysis of the shortcomings of the wavelet neural network (WNN) model, an improved WNN model to predict SCB is proposed herein. Th...

Background: In the pulmonary nodules computer aided diagnosis systems (CAD), feature selection plays an important role in reducing the false positive rate and improving the system accuracy. To solve the problem of feature selection techniques by which the diversity of features was damaged in the process of distinguishing malignant pulmonary nodules from benign pulmonary nodules, this study developed a novel feature selection algorithm for improving the accuracy of traditional computer-aided differential diagnosis for benign and malignant classification of pulmonary nodules. Method: Firstly, we divided the extracted features of nodules into several groups by using Gaussian mixture model (GMM). Secondly, we applied Relief and sequential forward selection (SFS) algorithm to find local optimum features dataset for each group. Afterwards, we used the optimumpath forest (OPF) classifier with the found features dataset to obtain the classification results. Finally, the local optimum features dataset with the highest area under curve AUC in all groups were added into the final selected set. Results: According to collected pulmonary nodules on computed tomography (CT) scans, tested with two set of samples, we achieved an average accuracy of 89.5%, sensitivity of 87.1% and specificity of 90.9% on the first set of samples, and 90.1%, 88.7% and 92.1% on the second set of samples. The areas under the receiver operating characteristic (ROC) curves based on these two sample sets were 95.2%, and 96.3% respectively. Conclusions: This study shows that the proposed method was promising for improving the pulmonary nodules computer aided diagnosis systems performance of benign and malignant pulmonary nodules.

Introduction: We investigated the impact of long-term levothyroxine (LT4) treatment on bone mineral density (BMD) and bone turnover markers (BTMs) in premenopausal women with differentiated thyroid cancer (DTC) after thyroidectomy. Material and methods: Sixty-five premenopausal women who received LT4 therapy at least one year (range, 1.5–9.0 years) after thyroidectomy for DTC and 50 premenopausal women without thyroid diseases were enrolled in this study. We measured the Z-scores of lumbar and hip BMD, serum free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), intact parathyroid hormone (iPTH), N-terminal propeptide of type 1 N procollagen (P1NP), C terminal telopeptide of type 1 collagen (CTX-1), calcium (Ca), phosphorus (P), vitamin D3, and alkaline phosphatase (ALP) in all participants. Results: In DTC subjects, serum TSH levels were lower, and serum FT4, P1NP, CTX-1, and ALP levels were higher compared with controls. The prevalence of low BMD was higher in DTC subjects than in controls. Multivariate logistic regression analysis showed that serum TSH levels were negatively associated with CTX-1 and ALP. Conclusions: We found a high prevalence of low BMD among premenopausal women who received long-term LT4 therapy for DTC after thyroidectomy. Long-term TSH suppression therapy was a significant risk factor for decreased bone strength, mainly by increasing bone turnover.

While the Internet provides convenience to people, security issues for data still exist. Images require huge storage space and cloud computing can help users to solve this problem, but the data on the cloud may be leaked. This paper proposes an image data processing scheme based on the Paillier encryption algorithm. To ensure the security of image, we use the Paillier algorithm to process the image data before uploading it to the cloud platform. Then, the cloud server applies the SIFT algorithm to extract and save the image features. A scheme for checking similarity is also proposed to ensure the security of the image in the cloud. We also discuss its application in outsourced image processing.

In view of the fact that most of the home planting information technology level is low, a home smart planting system is designed applying Internet of Things technology. The system uses STM32 microprocessor as the main control chip to design terminal nodes. Each node has the function of sensing the planting environment information such as air temperature and humidity, soil humidity, light intensity, and the control function of drip irrigation and light supplement switch. The number of terminal nodes can be increased or decreased according to the needs, which can meet the basic needs of home smart planting. The terminal node connects the home wireless router through WiFi module, and then transmits the sensing data to the cloud server. The user terminal is a smartphone. User can log in to the home smart planting system through the mobile APP, and then interact with the terminal device, monitor the sensing data in real time, view video of the plant growth status and planting environment, and control the drip irrigation and light supplement devices remotely. The automatic control mode can also be set to control the devices automatically. The system has the advantages of simple architecture, convenient deployment and use. The rented cloud server is used for data storage and data processing. Data is transmitted using the services provided by operators. It is safe and reliable with low operation and maintenance cost, and is suitable for home planting.

In recent years, deep-learning-based super-resolution (SR) methods have obtained impressive performance gains on synthetic clean datasets, but they fail to perform well in real-world scenarios due to insufficient real-world training data. To tackle this issue, we propose a conditional-normalizing-flow-based method named IDFlow for image degradation modeling that aims to generate various degraded low-resolution (LR) images for real-world SR model training. IDFlow takes image degradation modeling as a problem of learning a conditional probability distribution of LR images given the high-resolution (HR) ones, and learns the distribution from existing real-world SR datasets. It first decomposes the image degradation modeling into blur degradation modeling and real-world noise modeling. It then utilizes two multi-scale invertible networks to model these two steps, respectively. Before applied into real-world SR, IDFlow is first trained supervisedly on two real-world datasets RealSR and DPED with negative log-likelihood (NLL) loss. It is then used to generate a large number of HR-LR image pairs from an arbitrary HR image dataset for SR model training. Extensive experiments on IDFlow with RealSR and DPED are conducted, including evaluations on image degradation stochasticity, degradation modeling, and real-world super resolution. Two known SR models are trained with IDFlow and named as IDFlow-SR and IDFlow-GAN. Testing results on the RealSR and DPED testing dataset show that not only can IDFlow generate realistic degraded images close to real-world images, but it is also beneficial to real-world SR performance improvement. IDFlow-SR achieves 4× SR performance gains of 0.91 dB and 0.161 in terms of image quality assessment metrics PSNR and LPIPS. Moreover, IDFlow-GAN can super-resolve real-world images in the DPED testing dataset with richer textures and maintain clearer patterns without visible noises when compared with state-of-the-art SR methods. Quantitative and qualitative experimental results well demonstrate the effectiveness of the proposed IDFlow.

An accurate tensor estimation function is a guarantee for acquiring a reliable underground electrical structure model. The past studies regarding transmission functions aim at obtaining smooth and continuous sounding curves through the time domain by data screening means such as least square, robust and remote reference processing methods as well as digital filtering techniques such as wavelet denoising, empirical mode decomposition and Hilbert–Huang transform denoising. As SNRs in different field measurement sites are different, if different data processing algorithms like single-site processing, remote reference processing, magnetic field correlation-based remote reference processing and magnetic channel sharing are used for data processing at the same measuring site, 3N-2 groups of sounding curves with different qualities will be obtained. Manual screening method has been mainly used in the past for screening of the abovementioned 3N-2 groups of sounding curves. If the quantity of sounding site is large, this manual screening method is of low efficiency without quantitative screening indexes. In this paper, discrete Frechet distance is mainly utilized to carry out quantitative screening of 3N-2 groups of sounding curves processed through the abovementioned four processing methods in order to obtain smooth, continuous and reliable sounding curves, thus avoiding subjectivity and low efficiency problems in the past manual screening. A quantitative sorting of all kinds of data processing results can be conducted, and the optimal sounding curve is finally picked out for data inversion. The above sounding curve selection result has improved processing accuracy of magnetotelluric data.

The Micro-Resistive WELL ( μ RWELL) detector is a novel spark-protected Micro-Pattern Gaseous Detector (MPGD) with a single well-type amplification stage. By integrating a well-type amplification structure with a resistive electrode on a readout board ( μ RWELL PCB), the μ RWELL detector is only composed of a cathode PCB and a μ RWELL PCB, resulting in a very simple assembling process without any gluing, stretching, tensioning or other complicated operation commonly practiced in assemblies of other MPGDs. A critical component of the μ RWELL detector is the resistive electrode that is responsible for spark suppression, and hence maintains stable operation of the detector. In this paper, we present a 10 cm × 10 cm μ RWELL detector with two layers readout strips perpendicular to each other. Its resistive electrode is made of a thin coating of Diamond-Like Carbon by magnetron sputtering technique. The gain performance and rate capability of this μ RWELL detector were systematically studied with X-rays, and its detection efficiency and spatial resolution were measured with a test beam. The detector can achieve a spatial resolution better than 70 μ m in both dimensions while the detection efficiency is higher than 90%.

The rapid development of the construction industry in China has introduced unprecedented quality-related problems in the country’s building industry. In response to this issue, the government has established various complaint channels to report quality problems. Therefore, building quality complaints (BQCs) need to be classified and solved by respective agencies or departments rapidly for avoiding adverse impact on the safety, health, and well-being of people. However, the current process of classifying BQCs is labor intensive, time consuming, and error prone. An automatic complaint classification is required to improve the effectiveness and efficiency of complaint handling, but studies on this issue are limited. Prevailing text classification research in construction has focused on utilizing conventional shallow machine learning. By contrast, this study explores a novel convolutional neural network (CNN)-based approach that incorporates a deep-learning method to automatically classify the short texts contained within BQCs. The presented approach enables capturing the semantic features in BQC texts and automatic classification of the BQCs into predefined categories. After the model optimization, tests are conducted to examine the practical application of the text classification approach compared with Bayes-based and support vector machine classifiers. Results indicate that the developed CNN-based approach performs well in the Chinese BQC classification with limited manual intervention and few complicated feature engineering.

A new model for a multi-layer thermoacoustic transducer that takes full advantage of the excellent the thermoacoustic characteristics of carbon nanotube (CNT) thin film is developed. Analytical exact solutions of acoustic pressure distribution for both near and far fields are presented with engineering interpretation. By assuming acoustically transparent stacking layers and a total thickness less than a tenth of the sound wavelength, the acoustic pressure field relations between parallel layers are established that subsequently yield the sound pressure for the overall stacking films. A comparison between theoretical prediction and experiment for single-layer and multi-layer CNT thin films is presented and excellent agreement is reported. Key influencing factors including the total layer number, and the spacing between separated films and gas medium on the sound pressures level (SPL) are investigated and discussed in detail. Some design criteria that increases the SPL of the multi-layer CNT thin film transducer are also established.

During landfall of Typhoon Haikui in Eastern China in 2012, ground level wind data were recorded using a smart monitoring system installed on JiuBao Bridge in Hangzhou, China. This article documents the mean flow and turbulence characteristics from data recorded during the storm. The results show that both turbulence intensity and gust factor decrease with the increase in the mean wind velocity. However, as the mean wind velocity increases, this trend gradually attenuates. The peak factor distribution with gust averaging time duration derived with the Typhoon Haikui data agrees well with the Durst curve. However, the longitudinal gust factor derived from the typhoon wind-speed record in this study is higher compared with the curves proposed by Durst and Krayer-Marshall. Analyses of the gust factor distribution with the turbulence intensity during the passage of the storm reveal a similarity to the empirical curves of Ishizaki and Choi. Results show that the relationship between lateral turbulence and gust factors can be well represented by a quadratic polynomial. Turbulence scale increased with mean wind velocity. The values of autocorrelation coefficients in longitudinal direction are larger than those in lateral direction. There has no obvious dependency of cross-correlation coefficients with mean wind velocity. In general, the wind characteristics in this study are shown to be very similar to those of winds under normal circumstance.

Wireless energy replenishment using mobile wireless charging equipment (WCE) has been a hot topic recently in areas of wireless rechargeable sensor networks (WRSNs) and mobile computing. Previous studies of multi-node charging planning usually assumed that the WCE was equipped with sufficient energy to accomplish the charging task. However, the energy of the WCE is limited in practice. Therefore, a multi-node charging planning algorithm with energy-limited WCE in WRSNs is proposed in this paper. Moreover, three charging planning models and their corresponding charging planning algorithms are proposed based on the different real conditions of the WCE: sufficient moving energy and limited charging energy, limited moving energy and sufficient charging energy, as well as both limited moving energy and charging energy. Furthermore, a hybrid simulated annealing discrete fireworks algorithm (HSA-DFWA) is proposed due to the NP-Hard complexity of the charging planning problem. The experiments show that compared to DFWA, MM-ANT, E-greedy, and P-greedy algorithms, the HSA-DFWA outperforms these algorithms empirically.

A new model for a thermoacoustic transducer array with high acoustic beam directivity applying devised carbon nanotube (CNT) thin films is developed. Analytical exact solutions of acoustic pressure...

Simple Summary Forensic entomologists utilize sarcosaprophagous insect species to estimate the postmortem interval to aid death investigations. In this paper, we present the recent chronology of forensic entomology in China and illustrate how identification, development, and succession data are obtained and applied at the scale of such a large country. To overcome the difficulties and challenges forensic entomology faces in China, a number of countermeasures are provided. Abstract While the earliest record of forensic entomology originated in China, related research did not start in China until the 1990s. In this paper, we review the recent research progress on the species identification, temperature-dependent development, faunal succession, and entomological toxicology of sarcosaprophagous insects as well as common applications of forensic entomology in China. Furthermore, the difficulties and challenges forensic entomologists face in China are analyzed and possible countermeasures are presented.

Objective: To analyze the epidemiological characteristics of asymptomatic SARS-CoV-2 novel coronavirus infection cases in Yinchuan City, Ningxia. Methods: Cases of asymptomatic SARS-CoV-2 novel coronavirus infections discovered in Yinchuan City from February 2020 to April 2020 were taken as the research subjects, epidemiological and laboratory examination data were collected, and statistics and analysis were conducted. Results: Among the 10 cases of asymptomatic infection, 6 were males and 4 were females, with an average age of (34.3±2.1) years; 2 cases had a history of living in Hubei, 8 cases had a history of close contact with confirmed cases; 6 cases had no other underlying diseases while the other 4 cases were accompanied by one or more underlying diseases; 6 cases were diagnosed as asymptomatic infection by serum-specific IgM antibody detection; 4 cases were diagnosed as asymptomatic infection by throat swab testing; there were 4 cases with alanine base transferase level higher than the normal range, 2 cases with aspartate aminotransferase level higher than the normal range, and 3 cases with D-dimer level out of the normal range. Conclusion: The asymptomatic SARS-CoV-2 novel coronavirus infection has certain clinical characteristics; it is infectious and has insidious characteristics. Even if medical observation is dismissed, monitoring is still needed to prevent the pandemic from rebounding.

Document-level relation extraction (DocRE) models generally use graph networks to implicitly model the reasoning skill (i.e., pattern recognition, logical reasoning, coreference reasoning, etc.) related to the relation between one entity pair in a document. In this paper, we propose a novel discriminative reasoning framework to explicitly model the paths of these reasoning skills between each entity pair in this document. Thus, a discriminative reasoning network is designed to estimate the relation probability distribution of different reasoning paths based on the constructed graph and vectorized document contexts for each entity pair, thereby recognizing their relation. Experimental results show that our method outperforms the previous state-of-the-art performance on the large-scale DocRE dataset. The code is publicly available at https://github.com/xwjim/DRN., Comment: 11 pages, 4 figures, 5 tables. Findings of ACL 2021

As the demand for electricity continues to grow, the coverage of transmission lines is getting larger. Despite the continuous improvement of the grid system, transmission lines are still vulnerable to various natural disasters. At the same time, many transmission lines are installed in areas with harsh environments and other places that are difficult for operation and maintenance personnel to reach, which brings huge challenges to the operation and maintenance of transmission lines. Therefore, how to effectively detect the status of the transmission line and ensure the normal operation of the power grid has become an important research topic in the power system. The existing video surveillance-based methods need to decode the video, which has poor real-time performance. Therefore, this paper proposes a transmission line abnormality monitoring method based on video stream analysis. Using the extracted parameters for judgment before decoding the video stream can effectively improve the real-time performance of online monitoring of the transmission line and greatly shorten the time required for abnormal alarms.

Three-dimensional embedded devices are very good in application, but there are many failure cases. In this paper, we develop a set of DPA methods and procedures through failure cases in recent years.

To solve the problem of difficult modulation recognition of emitter signals at low signal-to-noise and low recognition rate caused by improper feature extraction, a new end-to-end modulation recognition algorithm based on deep learning is proposed. The algorithm can effectively recognize LFM, Costas, Frank code, and P1-P4 code. Firstly, an end-to-end modulation recognition network based on bi-directional long-short term memory is proposed to extract the time features of emitter signals for recognition. Then, to extract the space features of emitter signals, it explores a one-dimensional convolutional neural network structure. Finally, one-dimensional convolutional neural network and bi-directional long-short term memory are combined with pre-training and fine-tuning to accelerate the convergence speed and improve the recognition rate. The simulation results show that the recognition rate of the algorithm can reach 75.23% at the −6dB signal-to-noise ratio. It can also be used for modulation recognition in the mechanical field.

Recently Cyber-Physical-Human Systems (CPHS) have been attracted much attention. Unlike existing computing paradigms such as Cloud Computing and Edge Computing, CPHS usually comprise many types of hybrid resources such as software and hardware in the cloud layer, network devices and edge servers in the network layer, as well as the physical devices and human participation. Such hybrid resource makes it challenging to efficient resource scheduling for CPHS. Previous studies on CPHS often focus on improving scheduling efficiency for specific application scenarios and thus are limited for other CPHS. In this work, we present a unified hybrid resource scheduling framework for CPHS, including a hybrid resource model for multiple types of resource providers and participants in CPHS, a hybrid resource orchestration tool that can describe our hybrid resource model as the resource requirements, and a scheduling mechanism to satisfy the resource requirements especially by considering the resource constraints and crosslayer scheduling in CPHS. We perform the case study of typical CPHS scenarios and show that our hybrid resource model and scheduling framework is effective, and provide a new view for improving the resource provision efficiency in CPHS.

In order to address the problem that the geodesic distance of Lie Group machine learning method has failed to use the sample class information, this paper proposes an improved Lie Group machine learning algorithm. Firstly, this algorithm maps the vector space consisting of sample dataset into a differential manifold, and each sample corresponds to a point in the manifold structure; then, the differentiation value is used to replace the Euclidean distance for calculation of the geodesic distance between samples, and the class of test sample is determined according to the geodesic distance relation between the test sample and the sample in training dataset. The results of classification experiment prove that compared with the traditional Lie Group machine learning classification (LGC) algorithm, the K-nearest neighbor (KNN) algorithm and support vector machine (SVM) classification\algorithm, the improved classification algorithm has higher classification accuracy.

With the rapid development of the Internet, network technology has gradually penetrated into the lives of ordinary people, and the advent of the era of big data has made people finish work through computers. The development of cloud computing provides people a Virtual platform to solve the computing problems. However, when users outsource data to cloud computing platforms, cloud security issues also come behind. Because cloud computing technology is an emerging technology, technical confidentiality is not enough to meet user needs, and the relevant laws and regulations of the technology are not perfect. It also lacks perfect service agreement management standards and calculation mechanism. To solve the security problems of cloud computing, homomorphic encryption came into being. By analyzing the development trend of cloud computing and the application of homomorphic encryption in cloud computing, this paper introduces the basic knowledge of Paillier encryption algorithm, and proposes a cloud data secure outsourcing protocol, which partially solves the blind computing problems. The method, which transfers the data calculation to the two parties separately through key division, can ensure the security of the data in the calculation and protect the integrity and confidentiality of the outsourced data in the entire process. Through protocol analysis, this method can achieve the expected result.

The popularization and application of computers and the close integration of computer technology and communication technology make the transmission of information more convenient. In daily work, the use of file transfer systems can greatly improve work efficiency. While the file transfer system brings us convenience, it also has security risks. Through the research and analysis of the current situation of office informatization and information security issues, according to the daily work needs of enterprises and individuals, combined with file transfer technology and cryptography theory, a file transfer system with strong security and convenient operation is designed and implemented. The system consists of four modules: file transfer module, encryption and decryption module, instant communication module and log management module. It can complete the encrypted transmission of files, provide instant messaging functions, and keep transfer records. By testing and using the system, the secure transmission of data can be realized, which provides a good guarantee for the security of the transmitted files.

ZnS-CuInS2 (CIZS) solid solution is synthesized via a chitosan gel-assisted method. The bandgap of the obtained CIZS semiconductor is 1.68 eV, and the position of the flat potential is 0.86 eV. When Ru is selected as the cocatalyst, the photocatalytic hydrogen evolution rate of CIZS@Ru reaches 4.86 mmol g−1h−1 under visible-light (λ＞420 nm) radiation, and it presents good stability. Besides, CIZS@Ru also shows good CO2 reducibility, where the product contains hydrogen, methane, and a small amount of formate. The photoelectrochemical tests also suggest that Ru is the optimum cocatalyst compared with Rh, Pt and Ni. Meanwhile, the internal mechanism of photogenerated charge transfer between semiconductors and cocatalyst has been revealed by the band bending concept. This work provides a novel and facile strategy to fabricate p-type CIZS semiconductor with high visible-light photocatalytic reduction performance.

Laparoscopic common bile duct exploration (LCBDE) has been recently introduced for management of CBD stone in patients with previous biliary surgery history. The aim of this study was to evaluate the safety and effectiveness of primary closure in patients with previous biliary surgery history compared to T-tube drainage. Eighty patients with previous biliary surgery history including laparoscopic cholecystectomy, open cholecystectomy, or open common bile duct exploration were enrolled in the retrospective study. The patients were divided into 2 groups according to the methods of choledochotomy closure. Group A: patients with primary closure after LCBDE (n = 51); group B: patients with T-tube drainage after LCBDE (n = 29). Group A exhibited a shorter postoperative hospital stay and lower hospitalization expenses compared to group B. There was no significant difference in conversion rate to open surgery, operating time, intraoperative blood loss, bile leakage rate, overall complication rate, and stone recurrence rate between the 2 groups. Biliary stricture was not observed in the 2 groups during the follow-up period. Primary closure following LCBDE is safe and effective for the management of CBD stones in patients with previous biliary surgery history.

Real-time users cannot carry out real-time precise point positioning (RT-PPP) because they cannot receive real-time service (RTS) products from the international GPS service (IGS) in the case of interrupted communication. We address this issue by introducing a stable particle swarm optimized wavelet neural network (PSOWNN) to predict the short-term satellite clock bias in real time accurately. The predicted sequences of the new model are compared with those of the conventional linear polynomial, quadratic polynomial, gray system (GM (1,1)), and Kalman filter models. The results show that the accuracy of the proposed model is better than that of these four models. The average prediction accuracy of the 30-min and 60-min forecasting has improved by approximately (79.3, 82.4, 79.1, 97.4) % and (97.4, 82.9, 87.7, 98.9) % and is better than 0.3 ns during 30-min and 1-h forecasting. The RTS products can thus be replaced with the short-term clock products predicted by the PSOWNN model to meet the precision requirements of RT-PPP.

Objective: To explore the effect of respiratory rehabilitation training on the quality of life of pneumoconiosis patients. Methods: 76 pneumoconiosis patients who were treated in our hospital from April 2017 to December 2019 were selected as the research object, and randomly divided into 2 groups according to the order of admission by coin tossing, 38 cases in each group. The control group carried out health knowledge education on the basis of conventional treatment, and the observation group combined with respiratory rehabilitation training on the basis of the control group to compare the quality of life and lung function of the two groups of patients. Results: After 2 months of nursing care, scores of GQOLI-74 scale and pulmonary ventilation function indexes in the observation group were higher than those in the control group, with statistically significant differences (P