Search

In order to develop organic compounds for achieving highly efficient anticorrosion of mild steel in HCl solution, this study proposed to synthesize new dendritic molecules (DMs 1, 2) containing double norfloxacin skeletons. In addition, the linear molecule (LM) carrying a single norfloxacin framework was also prepared as a reference. The chemical structures of the studied molecules were fully characterized by nuclear magnetic resonance spectroscopy (NMR) (1D and 2D NMR spectroscopy, 1H, 13C, 19F), mass spectroscopy, Fourier-transform infrared spectroscopy. For this purpose, the adsorption of the studied molecules on mild steel was investigated by different means. Furthermore, the potential kinetic polarization and electrochemical impedance spectroscopy were used to survey the anticorrosion of the studied molecules in HCl solution at 298 K. It is shown that the DMs displayed superior corrosion inhibition effect on mild steel over the LM in acid medium at 298 K (the maximal corrosion inhibition efficiency, LM, 87.80%, DM1, 96.00%, DM2, 96.26% at 0.015 mM). The anticorrosion and adsorption mechanisms of the studied molecules for mild steel were further understood by material simulation and adsorption isotherms.

The segmentation of retinal vessels plays an essential role in the early recognition of ophthalmic diseases in clinics. Increasingly, approaches based on deep learning have been pushing vessel segmentation performance, yet it is still a challenging problem due to the complex structure of retinal vessels and the lack of precisely labeled samples. In this paper, we propose a self-paced dual-axis attention fusion network (SPDAA-Net). Firstly, a self-paced learning mechanism using a query-by-committee algorithm is designed to guide the model to learn from easy to hard, which makes model training more intelligent. Secondly, during fusing of multi-scale features, a dual-axis attention mechanism composed of height and width attention is developed to perceive the object, which brings in long-range dependencies while reducing computation complexity. Furthermore, CutMix data augmentation is applied to increase the generalization of the model, enhance the recognition ability of global and local features, and ultimately boost accuracy. We implement comprehensive experiments validating that our SPDAA-Net obtains remarkable performance on both the public DRIVE and CHASE-DB1 datasets.

Based on the team faultlines theory, this paper examines the impact of different types of board faultlines on innovation in cross-border mergers and acquisitions (M&As). Using the data of Chinese firms’ cross-border M&As, this paper finds that information-based faultlines and social category faultlines have an inverted U-shaped relationship with innovation in cross-border M&As. Cultural distance weakens the role of information-based faultlines on innovation in cross-border M&As. This paper enriches the research on the factors influencing innovation in cross-border M&As and provides a basis for firms to select board members and create suitable board faultlines.

BHACS was established to fill the lack of a large representative cohort of the elderly based on general populations and was designed to evaluate the prevalence, incidence, and natural history of cognitive decline, functional disability, and conventional vascular risk factors. The objectives were to determine the evolution of those diseases by estimating rates and impact factors of progression and reversion to adverse outcomes including dementia, cardiovascular events, cancer, and all-cause death. Therefore, it can provide evidence to help policy-makers develop better policies to promote healthy aging in China. BHACS was formed by three cohorts in Beijing with a total population of 11 235 (6281 in urban, 4954 in rural), the baseline survey time and follow-up time differed between the three. BLSA was followed up in 2014 with the baseline survey completed in 2009; CCHS-Beijing was followed up twice in 2015-2016 and 2018-2019 as its baseline survey was performed in 2013-2015; BECHCS was followed up every two or three years with the baseline investigation completed in 2010-2014. Data were collected using questionnaires, physical measurements, and laboratory analysis. Topics of BHACS include numerous physical and mental health indicators, living habits, and personal, familial, and socio-economic health determinants. There is no immediate plan to make the cohort data freely available in the public domain, specific proposals for further collaboration are welcome.

Zero-shot learning (ZSL) is implemented by transferring knowledge from seen classes to unseen classes through embedding space or feature generation. However, the embedding-based method has a hubness problem, and the generation-based method may contain considerable bias. To solve these problems, a joint model with multiple generative adversarial networks (JG-ZSL) is proposed in this paper. Firstly, we combined the generation-based model and the embedding-based model to build a hybrid ZSL framework by mapping the real samples and the synthetic samples into the embedding space for classification, which alleviates the problem of data imbalance effectively. Secondly, based on the original generation-method model, a coupled GAN is introduced to generate semantic embeddings, which can generate semantic vectors for unseen classes in embedded space to alleviate the bias of mapping results. Finally, semantic-relevant self-adaptive margin center loss was used, which can explicitly encourage intra-class compactness and inter-class separability, and it can also guide coupled GAN to generate discriminative and representative semantic features. All the experiments on the four standard datasets (CUB, AWA1, AWA2, SUN) show that the proposed method is effective.

3D object detection and recognition are crucial tasks for many spatiotemporal processing applications, such as computer-aided diagnosis and autonomous driving. Although prevalent 3D Convolution Nets (ConvNets) have continued to improve the accuracy and sensitivity, excessive computing resources are required. In this paper, we propose Leaky Integrate and Fire Networks (LIF-Nets) for 3D detection and recognition tasks. LIF-Nets have rich inter-frame sensing capability brought from membrane potentials, and low power event-driven mechanism, which make them excel in 3D processing and save computational cost at the same time. We also develop ResLIF Blocks to solve the degradation problem of deep LIF-Nets, and employ U-LIF structure to improve the feature representation capability. As a result, we carry out experiments on the LUng Nodule Analysis 2016 (LUNA16) public dataset for chest CT automated analysis and conclude that the LIF-Nets achieve 94.6% detection sensitivity at 8 False Positives per scan and 94.14% classification accuracy while the LIF-detection net reduces 65.45% multiplication operations, 65.12% addition operations, and 65.32% network parameters. The results show that LIF-Nets have extraordinary time-efficient and energy-saving performance while achieving comparable accuracy.

Most of the existing deep learning algorithms are supervised learning and rely on a tremendous number of manually labeled samples. However, in most domains, due to the scarcity of samples or the excessive cost of labeling, it would be impracticable to provide numerous labeled training samples to the network. In this paper, a few-shot video classification network termed Hierarchical Motion Excitation Network (HME-Net) is proposed from the perspective of accumulated feature-level motion information. An HME module composed of Motion Excitation (ME) and Interval Frame Motion Excitation (IFME) is designed to extract feature-level motion patterns from adjacent frames and interval frames. The HME module can discover and enhance the feature-level motion-sensitive information in the original features. The accumulative time window is expanded to four frames in a hierarchical manner, which achieves the purpose of increasing the receptive field. After extensive experimentation, HME-Net is demonstrated to be able to consistently outperform the existing few-shot video classification models. On the UCF101 and HMDB51 datasets, our method is established as a new state-of-the-art technique for the few-shot settings of five-way three-shot and five-way five-shot video recognition.

Pulmonary nodule detection is essential to reduce the mortality of lung cancer. One-stage detection methods have recently emerged as high-performance and lower-power alternatives to two-stage lung nodule detection methods. However, it is difficult for existing one-stage detection networks to balance sensitivity and specificity. In this paper, we propose an end-to-end detection mechanism combined with a channel-wise attention mechanism based on a 3D U-shaped residual network. First, an improved attention gate (AG) is introduced to reduce the false positive rate by employing critical feature dimensions at skip connections for feature propagation. Second, a channel interaction unit (CIU) is designed before the detection head to further improve detection sensitivity. Furthermore, the gradient harmonizing mechanism (GHM) loss function is adopted to solve the problem caused by the imbalance of positive and negative samples. We conducted experiments on the LUNA16 dataset and achieved a performance with a competition performance metric (CPM) score of 89.5% and sensitivity of 95%. The proposed method outperforms existing models in terms of sensitivity and specificity while maintaining the attractiveness of being lightweight, making it suitable for automatic lung nodule detection.

This study reported on a variety of target phenanthridine dyes carrying hydroxy groups that were employed as near-IR two-photon absorption (TPA) photostabilizers for polymers. It was demonstrated that the target phenanthridine dyes underwent reversible excited state proton transfer (ESPT) in strong polar protic and aprotic solvents, such as ethanol, which was more remarkable under near-IR two-photon excitation. It was determined that the target dyes processed regular molecular stacking in mixed organic solvents/H2O solvents (such as ethanol/H2O, 90% water volume). The orderly molecular aggregation of the target phenanthridine dyes intensified the reversible ESPT emission under one-photon and near-IR two-photon excitation, respectively. Hence, the absorbed energy was consumed by the reversible ESPT process of the target dyes. As a consequence, the target dyes were evaluated as near-IR TPA photostabilizers for polymers. In contrast, the reference phenanthridine dyes lacking of hydroxy groups did not exhibit ESPT or photostability under light irradiation.

In this study, three new gemini ionic liquids (GILs 1–3) were designed and prepared. The results showed that the GILs could undergo orderly molecular aggregation in sulfuric acid solution, which reached a steady aggregation state at 0.050 mM with 8 h evolving time. The tough chemical adsorption of the GILs aggregates on copper surface was confirmed by different means such as X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD), which could be formed by the Cu(I)-GILs coordination complexes. The yielded hydrophobic self-assembled monolayers (SAMs) adsorption of the GILs aggregates on copper surface were demonstrated by the measurements of spectroscopic ellipsometry and contacting angles (such as the GIL 3, the adsorption layer thickness, 5.70 nm, the contacting angle, 124.5 °). The copper corrosion inhibition effects of the GILs in sulfuric acid solution were measured by the electrochemistry analysis. The obtained results show the GIL 3 achieved over 96% corrosion inhibition efficiency at 0.050 mM based on electrochemical impedance spectroscopy (EIS). The decrease of charge transfer as well as the increase of double layer capacitance may make major contributions in copper corrosion resistance. The adsorption behavior of the GILs aggregates on the Cu surface was found to obey the Langmuir isotherm plots. In further consideration of the other advantages of the target GILs such as water solubility, negligible toxicity as well as easy preparation based on commercial starting materials, a large application potential could be expected.

This study proposes to use molecular aggregation method of organic ionic liquid molecules (OILMs) to intensify adsorption and corrosion inhibition to copper surface in sulfuric acid solution. Three new single imidazole-type OILMs including long carbon chain were synthesized, which were characterized by different methods such as nuclear magnetic resonance (NMR) and elemental analysis. The results suggest that the OILMs could process regular molecular assembly in sulfuric acid solution. The formed OILMs aggregates have been shown dependence on the OILMs concentrations and aggregation time. The chemistry adsorption of the OILMs aggregates on copper surface was demonstrated by various means including attenuated total reflection infrared spectroscopy (ATR-IR), X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), and spectroscopic ellipsometry and contacting angles, and atomic force microscopy (AFM) as well as scanning electron microscopy (SEM) imaging. Thus, the hydrophobic adsorption layers of the OILMs aggregates were yielded on copper surface. The electrochemistry survey suggests that the OILMs aggregates present nice corrosion resistance effect to copper in sulfuric acid solution (the maximal corrosion inhibition efficiency, >93%). The results were further understood by the theoretical simulation computation.

One of the main problems for safe autonomous driving vehicles that have not been solved completely is the high-precision and timely lane detection. In this work, we present a novel operator for railway detection to settle these tasks based on lane detection for the first time, called adaptive multi-scale fusion Sobel operators. The new operators can eliminate the noises generated by the environment in the railway image and derive more integrated edge feature information from the 0°, 45°, 90°, and 135° detection via 4 matrixes of 3 * 3 operators for permutation and summation. The image processing for railway detection includes the preprocess for images, railway edge detection, and track line polynomial fitting. Our experiment has validated that this improved detection method has realized the high accuracy and efficiency for rail detection. The dynamic rail detection and identification in the video of the railway track prove that this method has a significant effect on the left and right curved railway detection. It has good robustness and applicability.

Since morphology of retinal blood vessels plays a key role in ophthalmological disease diagnosis, the automatic retinal blood segmentation method is essential for computer-aided diagnosis system. In this paper, a supervised method which is based on multi-level convolutional neural network is proposed to separate blood vessels from fundus image. By using both local and global feature extractors, small vessels can be well distinguished and global spatial consistency of the image can be ensured. Meanwhile, unsupervised pre-processing and postprocessing methods are applied to achieve better segmentation results. Experiment results on public database show that the proposed method outperforms the state-of-the-art performance (AUC up to >0.978) on DRIVE database.

Oil is essential to our life mainly in transportation, and thus the productivity of oil well is very important. Classification of oil wells can make it easier to manage wells to ensure good oil productivity. Machine learning is an emerging technology of analyzing data in which cluster is a good way to do classification. The paper will apply two kinds of cluster method to the data from Dagang oil well and then do analysis on not only the classification results but also the choice of method for future analysis.

Dynamic early warning makes great sense for oil management to keep safety and stability of oil production. In this paper, we derive the production regression model, predict production with 10 oil parameters based on Least Absolute Shrinkage and Selection Operator (Lasso) and Least Angle Regression (LARS) methods. The 10 most relevant oil parameters are decided by the warning parameters selection method from kinds of different parameters, which makes the prediction more reliable. The accuracy of regression model achieves 97%. Then we get the warning thresholds based on 6σ. Oil parameters for warning threshold partition experiment are from the database of Tianjin oilfield. The experiment results show that our method is capable of warning both mild and severe situation, and the accuracy is 95%, which runs ahead in oil industry and has great popularization value.

In this paper, a robust blind extracting audio watermarking scheme based on Improved Spread Spectrum (ISS) and Quadrature Phase Shift Keying (QPSK) is proposed. The frequency of direct spread spectrum sequence signal is reduced to limit its bandwidth, and then modulated to a higher frequency band with QPSK so as to lower the influence of watermarking to listener, meanwhile improve signal-noise ratio (SNR). The amplitude of watermarking to be embedded is determined by the correlation of spread spectrum (SS) sequence and the high frequency part of host audio signal. Experimental results show that the proposed scheme is more robust and transparent than traditional ISS watermarking.