Search

Drawing on the international relations literature, the purpose of this paper is to investigate whether and when diplomatic relations between countries have an influence on firm internationalization speed. We discover that the friendlier bilateral diplomatic relations between home and host countries, the more beneficial it is to enhance firm internationalization speed. In addition, the higher trade openness of the host country, the smaller contribution of bilateral diplomatic relations for firm internationalization speed in the host country and friendly bilateral diplomatic relations have a stronger positive impact on internationalization speed when the internationalizing home enterprise is a state-owned enterprise (SOE). Our findings not only contribute to both the firm internationalization speed literature and the international relations literature, but also have managerial significance for how enterprises can use the friendly diplomatic relations between countries to improve firm internationalization speed. Keywords Diplomatic relations * Trade openness * Nature of firm ownership * Internationalization speed * International relations JEL Classification F23 * F55, 1 Introduction Internationalization speed, defined as the average internationalization state achieved by a firm over a specific period (Casillas & Acedo, 2013; Chetty et al., 2014), has become a core [...]

A heat-driven catch-and-release strategy for CoCl2capture is described. It is based on the use of an immobilized neutral dicyclohexylacetamide-based receptor Lsupported on polystyrene (PS-L). An X-ray diffraction analysis of a single crystal of L·CoCl2revealed an ion-pair complex comprising a hexacoordinated cobalt cation [L·Co]2+and a tetrachlorocobaltate anion [CoCl4]2–. Temperature dependent binding was seen, as inferred from UV–vis spectroscopic studies. Fits to the van’t Hoff equation yielded values of ΔH° = 12.4 kJ/mol and ΔS° = 56.0 J/K·mol for L+ CoCl2, and ΔH° = 16.5 kJ/mol and ΔS° = 85.0 J/K·mol for PS-L+ CoCl2in 95% ethanol. Consequently, cobalt capture and release are mediated by heating and cooling, respectively. The material PS-Lexhibits a preference for binding cobalt over manganese and nickel as inferred from Langmuir–Freundlich isotherm analyses that revealed binding constants of KLF= 88.5 M–1for CoCl2, 52.7 M–1for MnCl2, and 49.7 M–1for NiCl2. In a simulated ion mixture containing equimolar CoCl2, MnCl2, and NiCl2, ICP-MS analyses served to confirm that cobalt was selectively enriched to 52 mol % (from an initial level of ca. 32 mol %) after one catch-and-release cycle and 76.6% after three cycles. Our experimental results were validated by density functional theory calculations, which also show stronger binding of Co over Mn and Ni to L.

The dynamics of ion transport at the interface is the critical factor for determining the performance of an electrochemical energy storage device. While practical applications are realized in concentrated electrolytes and nanopores, there is a limited understanding of their ion dynamic features. Herein, we studied the interfacial ion dynamics in room-temperature ionic liquids by transient single-particle imaging with microsecond-scale resolution. We observed slowed-down dynamics at lower potential while acceleration was observed at higher potential. Combined with simulation, we found that the microstructure evolution of the electric double layer (EDL) results in potential-dependent kinetics. Then, we established a correspondence between the ion dynamics and interfacial ion composition. Besides, the ordered ion orientation within EDL is also an essential factor for accelerating interfacial ion transport. These results inspire us with a new possibility to optimize electrochemical energy storage through the good control of the rational design of the interfacial ion structures.

The electrical machines with fractional slot windings have been widely investigated. The main challenge of this type of machines is that the magnetomotive force (MMF) harmonics of armature winding is rich, which would lead to torque ripple, eddy current loss, and other adverse effects. In this article, multiple three-phase star–delta windings connection for multiphase electrical machine is proposed, which can efficiently reduce the MMF harmonics of armature winding. First, the general principle of harmonic elimination of multiphase electrical machine with multiple star—delta winding connection is deduced. Then, the influence of multiple star–delta windings of machines with different slot/pole combinations on MMF harmonics of armature winding is investigated. To verify the theoretical analyses, the models of permanent magnetic machine with 24-slot and 36-slot are established, and the multiple star–delta windings is compared with the conventional three-phase, star–delta windings and multiple three-phase windings in terms of torque, torque ripple, loss, vibration, and so on, so as to evaluate the superiority of this connection method. Finally, a prototype with 24-slot/10-pole is manufactured and tested. The test results verify the rationality of the theoretical analyses and simulations.

In this article, the offline fault estimation via a sampled-output-data-driven adaptive observer is investigated for a class of Takagi–Sugeno fuzzy descriptor nonlinear systems. The additive actuator faults and disturbances are involved. At the signal receiving end, the discrete sampled output signals from systems are obtained through wireless transmission. To achieve the offline fault estimation for the fuzzy descriptor system, the cubic spline interpolation fitting approach is first applied to fit the receiving sampled output data into a dynamic fitting function. Then, the dynamic fitting function is further applied to construct and drive an adaptive observer related to the system in consideration. Compared with the existing observers, the designed discrete sampled-output-data-driven adaptive observer does not require interface docking with the original system equipment. The fault on the system can be estimated offline with a certain degree of accuracy via the provided observer. Stability conditions for the estimation error dynamics are given, which can be converted into linear matrix inequalities. Finally, two numerical models are used to validate the effectiveness of the proposed offline fault estimation method.

The conventional model predictive torque control (MPTC) used for dual three-phase permanent magnet synchronous motors (DTP-PMSMs) faces several challenges, such as significant computational burden, large current harmonics, and large torque and flux ripples due to the application of imprecise voltage vectors. To address these issues, a novel three-vector-based MPTC strategy was introduced in this article for DTP-PMSMs. First, virtual vectors were adopted to reduce the current harmonic components. Second, four candidate vector groups were determined based on torque deviation and flux position. Each candidate vector group comprised two virtual vectors and one zero vector, which enabled the precise control of torque and flux through the duty-ratio modulation strategy. Third, the switching sequence was optimized to facilitate implementation in digital controllers. Then, a cost function was defined to evaluate the average torque and flux ripples of each candidate vector group in a sampling period, and the one with the minimum cost function was applied in the next sampling period. This approach effectively reduced the current harmonics, torque ripple, and flux ripple while avoiding large computational burden. Finally, experiments were conducted to demonstrate the effectiveness of the proposed method.

The clinical outcomes between left-sided colon cancer and middle/low rectal cancer seem to be different. This study aimed to examine the effect of primary tumor location regarding the left-sided colon and middle/low rectum on the overall survival (OS) of patients who underwent colorectal hepatic metastasectomy.

A high-nuclear Co-added polyoxometalate (CoAP) was synthesized via a hydrothermal reaction: H14.5K9Na7.5–{[Co8(μ2–OH)(μ3–OH)2(H2O)2(Co(H2O)GeW6O26)(B-α-GeW9O34)2][BO(OH)2][Co12(μ2–OH)(μ3–OH)5(H2O)3(Co(H2O)GeW6O26)(GeW6O26)(B-α-GeW9O34)]}·46H2O (1). The polyoxoanion of 1contains a large Co20cluster gathered by lacunary GeW6O26and GeW9O34subunits. 1represents a one-dimensional (1D) chain formed by adjacent polyoxoanions coupling through a CoO6double bridge, showing the first example of a high-nuclear CoAP-based inorganic chain. 1served as an efficient electrocatalyst in oxygen evolution reactions (OERs).

Due to the flexibility and economy, energy storage systems (ESSs) have been widely deployed in wind energy power plants for power regulation. The coordination of generators and ESSs is critical for the wind farm (WF) operation. In this article, a two-stage voltage control is proposed in WFs equipped with distributed ESSs to achieve a near-global optimal performance without any centralized computations and communication links. To achieve it, the WF control is divided into two regions based on the demand-supply relationship of reactive power to consider the different operating conditions and multiple optimization objectives. The distributed battery-supercapacitor hybrid ESSs are deployed in each wind turbine within the WF to regulate the power output while ensuring voltage stability and enhancing the Var support capability. An optimization problem is formulated considering the node voltage deviation, the state of charge, and the cost of the storage units, which can be simply calculated using only local measurements through a decoupling method based on network topology characteristics. Further, an asynchronous response framework is designed for the battery and the supercapacitor to achieve a better economic benefit. Case studies in MATLAB/Simulink verify the effectiveness and reliability of the proposed method.

Multi-label image classification relies on a large-scale, well-maintained dataset, which may easily be mislabeled due to various subjective reasons. Existing methods for coping with noise usually focus on improving the model robustness in the case of single-label noise. However, compared with noisy single-label learning, noisy multi-label learning is more practical and challenging. To reduce the negative impact of noisy multi-annotations, we propose a universal approach for noisy multi-label learning (UNM). In UNM, we propose the label-wise embedding network which investigates the semantic alignment between label embeddings and their corresponding output features to learn robust feature representations. Meanwhile, mining the co-occurrence of multi-labels is also added to regularize the noisy network predictions. We cyclically change the fitting status of our label-wise embedding network to distinguish the noisy samples and generate pseudo labels for them. As a result, UNM provides an effective way to exploit the label-wise features and semantic label embeddings in noisy scenarios. To verify the generalizability of our method, we also test our method on Partial Multi-label Learning (PML) and Multi-label Learning with Missing Labels (MLML). Extensive experiments on benchmark datasets including Microsoft COCO, Pascal VOC, and Visual Genome explicitly validate the proposed method.

Titanium (Ti) and its alloys are widely used as hard tissue substitutes in dentistry and orthopedics, but their low bioactivity leads to undesirable osseointegration defects in the early osteogenic phase. Surface modification is an important approach to overcome these problems. In the present study, novel magnesium phosphate (MgP) coatings with controllable structures were fabricated on the surface of Ti using the phosphate chemical conversion (PCC) method. The effects of the microstructure on the physicochemical and biological properties of the coatings on Ti were researched. The results indicated that accelerators in PCC solution were important factors affecting the microstructure and properties of the MgP coatings. In addition, the coated Ti exhibited excellent hydrophilicity, high bonding strength, and good corrosion resistance. Moreover, the biological results showed that the MgP coatings could improve the spread, proliferation, and osteogenic differentiation of mouse osteoblast cells (MC3T3-E1) and vascular differentiation of human umbilical vein endothelial cells (HUVECs), indicating that the coated Ti samples had a great effect on promoting osteogenesis and angiogenesis. Overall, this study provided a new research idea for the surface modification of conventional Ti to enhance osteogenesis and angiogenesis in different bone types for potential biomedical applications.

Bispecific antibodies (BsAbs) are undergoing continued development for applications in oncology and autoimmune diseases. While increasing activity by having more than one targeting arm, most BsAb engineering employs single Fc engagement as monoclonal antibodies. Here, we designed a novel immunoglobulin gamma-1 (IgG1)-derived dual-Fc BsAb containing two Fc regions and two distinct asymmetric antigen binding arms comprising a Fab arm and another VHH domain. In conjunction with the knob-into-hole technology, dual-Fc BsAbs could be produced with a high yield and good stability. We explore how Fc engineering effects on dual-Fc constructs could boost the desired therapeutic efficacy. This new format enabled simultaneous bispecific binding to corresponding antigens. Furthermore, compared to the one-Fc control molecules, dual-Fc BsAbs were shown to increase the avidity-based binding to FcγRs to result in higher ADCC and ADCP activities by potent avidity via binding to two antigens and Fc receptors. Overall, this novel BsAb format with enhanced effector functionalities provides a new option for antibody-based immunotherapy.

Building change detection is critical for urban management. Deep learning methods are more discriminatory and learnable than traditional change detection methods. But in complicated backdrop environments, it is still difficult to precisely pinpoint change zones of interest. Most change detection networks suffer from inaccurate feature characterization during feature extraction and fusion. As a solution to these problems, we propose the use of multilevel feature fusion in conjunction with aware networks to detect building changes. To obtain multi-scale change characteristics, our Context-awareness network employs multi-scale patch embedding. Followed by multi-path Transformers to enhance learning and extract more suitable features. The multi-scale fusion module can ensure semantic consistency of change features, making detected change regions more accurate. Visual comparisons and quantitative evaluations of our method showed that it outperformed seven popular change detection methods on the LEVIR-CD dataset.

Bronchogenic cysts are rare congenital abnormalities, and usually asymptomatic until adulthood. We present a fetus prenatally diagnosed with a bronchogenic cyst, experiencing compression symptoms immediately after birth and underwent thoracoscopic surgery at 14 days old.

Electronic waste recycling is a recognized global challenge that requires new strategies to bind and release critical materials selectively, such as cobalt present in lithium-ion batteries. To address this challenge, hierarchical 3D-printed porous polymer scaffolds bearing supramolecular receptors were prepared using vat photopolymerization and their cobalt binding profiles were examined as a function of matrix polarity. By combining high-resolution digital light processing (DLP) with polymerization-induced phase separation (PIPS), functional acrylic copolymer networks with micrometer-level precision of geometry and nanometer-level pores were generated. Covalent integration of a methacrylate-functionalized bisdicyclohexyl acetamide (BDCA-MA) receptor enabled binding and release of cobalt(II) chloride (CoCl2) via a solvent polarity switch mechanism involving a change in solvent from ethanol to water. The present structures proved reusable as shown by sustained high binding efficiency over five bind and release cycles. This platform represents a “green” and energy conscious method for future electronic waste recycling.

In this paper, the molecular structures of vacuum residue (VR), coal liquefaction pitch (CLP), and coal tar pitch (CTP) are presented in detail, in order to explore their connection with mesophase properties. It is found that the H/C atom ratio of VR is as high as 1.32 because of abundant aliphatic structures. Its average aromatic ring number (RA) is only 3.88, and some of its internal aromatic rings are connected by bridged bonds. The mesophase pitch from VR (MP-VR) has a low mesophase content of 28.28–40.7% and is mostly of fiber texture. CTP has a degree of aromaticity (fa) of 0.90 and an RAof 10.65. The polycyclic aromatic molecule has only a few alkyl substituents at the periphery, resulting in a low H/C atom ratio of 0.56. The mesophase pitch from CTP (MP-CTP) has a mesophase content of 31.5–53.9% and has a distinct mosaic texture. CLP has the largest average molecular structure unit size, with an RAof 12.33 and a large degree of condensation. It contained certain cycloalkane structure and methyl substituents. The mesophase pitch from CLP (MP-CLP) has a mesophase content of 43.3–54.4% and contains abundant fiber and flaky textures. The regular and compact lamellar arrangement of planar aromatic macromolecules gives MP-CLP better graphitization properties, thermal stability, and rheological properties than MP-VR and MP-CTP.

Pavement distress classification is crucial for the maintenance of highways. Although many methods for classifying pavement distress are available, they all assume that training and testing datasets are drawn from the same distribution. When we introduce a new unlabeled dataset with a different distribution, the performance of existing methods decreases considerably due to domain shift, motivating us to look beyond the supervised setting to utilize unlabeled datasets directly in training a model. Therefore, we develop a novel unsupervised domain adaptation (UDA) framework, namely, the Self-supervised Adversarial Network (SSAN) for the first time in this study to conduct multi-category pavement distress classification on an unlabeled target domain. In particular, SSAN leverages adversarial domain adaptation (ADA) thoughts to align the features of different domains. However, distress typically occupies a small Section of high-resolution pavement images. Consequently, aligning features directly is unreasonable because the aligning procedure is still dominated by background features instead of foreground features, which are the most useful information for classification. Therefore, we design a pretext module, called Self-supervised Learning for the Target domain (SLT), to mine foreground information. To validate our method, we use two challenging pavement crack datasets, namely, the Chonqing University Bituminous Pavement Disease Detection (CQU-BPDD) and the Chongqing University Bituminous Pavement Multi-label Disease Detection (CQU-BPMDD) datasets. Moreover, extensive experiments demonstrate that SSAN outperforms state-of-the-art UDA methods.

Face reenactment aims to generate the talking face images of a target person given by a face image of source person. It is crucial to learn latent disentanglement to tackle such a challenging task through domain mapping between source and target images. The attributes or talking features due to domains or conditions become adjustable to generate target images from source images. This article presents an information-theoretic attribute factorization (AF) where the mixed features are disentangled for flow-based face reenactment. The latent variables with flow model are factorized into the attribute-relevant and attribute-irrelevant components without the need of the paired face images. In particular, the domain knowledge is learned to provide the condition to identify the talking attributes from real face images. The AF is guided in accordance with multiple losses for source structure, target structure, random-pair reconstruction, and sequential classification. The random-pair reconstruction loss is calculated by means of exchanging the attribute-relevant components within a sequence of face images. In addition, a new mutual information flow is constructed for disentanglement toward domain mapping, condition irrelevance, and condition relevance. The disentangled features are learned and controlled to generate image sequence with meaningful interpretation. Experiments on mouth reenactment illustrate the merit of individual and hybrid models for conditional generation and mapping based on the informative AF.

The dysregulation of long non-coding RNAs (lncRNAs) are involved in regulating tumor progression in multiple manner. However, little is known about whether lncRNA is involved in the translation regulation of proteins. Here, we identified that the suppressor of inflammatory macrophage apoptosis lncRNA (SIMALR) was highly expressed in nasopharyngeal carcinoma (NPC) tissues by analyzing the lncRNA microarray. Clinically, the high expression of SIMALRserved as an independent predictor for inferior prognosis in NPC patients. SIMALRfunctioned as an oncogenic lncRNA that promoted the proliferation and metastasis of NPC cells in vitro and in vivo. Mechanistically, SIMALRserved as a critical accelerator of protein synthesis by binding to eEF1A2 (eukaryotic translation elongation factor 1 alpha 2), one of the most crucial regulators in the translation machinery of the eukaryotic cells, and enhancing its endogenous GTPase activity. Furthermore, SIMALRmediated the activation of eEF1A2 phosphorylation to accelerate the translation of ITGB4/ITGA6, ultimately promoting the malignant phenotype of NPC cells. In addition, N-acetyltransferase 10 (NAT10) enhanced the stability of SIMALRand caused its overexpression in NPC through the N4-acetylcytidine (ac4C) modification. In sum, our results illustrate SIMALRfunctions as an accelerator for protein translation and highlight the oncogenic role of NAT10-SIMALR-eEF1A2-ITGB4/6 axis in NPC.

Service quality regulation for wind farms (WFs) can improve power generation and reduce operation and maintenance costs. In this paper, a hierarchical service quality optimal regulation method was proposed to achieve optimal system performance while proactively suppressing potential wind turbine (WT) failures. In the upper-lever, a global optimization problem was formulated based on model predictive control to meet the power commands of the transmission system operator and ensure node voltage safety. Then, the optimal power reference solved by the system-level optimization problem was given as input to the lower-level controller. In the lower-lever, tower displacement and converter temperature associated with power output were modeled as WT health indicators and integrated into a multi-objective optimization problem such that the optimization problem could improve the power flow while proactively suppressing WT failures. Thus, the traditional quality index was extended from individual WT health to system-level performance for the strategic generation of WF with higher reliability and improved economics. Case studies in MATLAB/Simulink demonstrated the effectiveness of the proposed method.

In this paper, we study the partial multi-label (PML) image classification problem, where each image is annotated with a candidate label set consisting of multiple relevant labels and other noisy labels. Existing PML methods typically design a disambiguation strategy to filter out noisy labels by utilizing prior knowledge with extra assumptions, which unfortunately is unavailable in many real tasks. Furthermore, because the objective function for disambiguation is usually elaborately designed on the whole training set, it can hardly be optimized in a deep model with stochastic gradient descent (SGD) on mini-batches. In this paper, for the first time, we propose a deep model for PML to enhance the representation and discrimination ability. On the one hand, we propose a novel curriculum-based disambiguation strategy to progressively identify ground-truth labels by incorporating the varied difficulties of different classes. On the other hand, consistency regularization is introduced for model training to balance fitting identified easy labels and exploiting potential relevant labels. Extensive experimental results on the commonly used benchmark datasets show that the proposed method significantly outperforms the SOTA methods.

Background: Although quercetin exhibits promising anti-tumor properties, its clinical application is limited due to inherent defects and a lack of tumor targeting.Objectives: This study aimed to prepare and characterize active targeting folate-chitosan modified quercetin liposomes (FA-CS-QUE-Lip), and its antitumor activity in vitroand in vivowas also studied.Materials and Methods: Box-Behnken Design (BBD) response surface method was used to select the optimal formulation of quercetin liposomes (QUE-LP). On this basis, FA-CS-QUE-LP was obtained by connecting folic acid chitosan complex (FA-CS) and QUE-LP. The release characteristics in vitroof QUE-LP and FA-CS-QUE-LP were studied. Its inhibitory effects on HepG2 cells were studied by the MTT method. The pharmacokinetics and pharmacodynamics in vivo were studied in healthy Wistar mice and S180 tumor-bearing mice, respectively.Results: The average particle size, zeta potential and encapsulation efficiency of FA-CS-QUELP were 261.6 ± 8.5 nm, 22.3 ± 1.7 mV, and 98.63 ± 1.28 %, respectively. FA-CS-QUE-LP had a sustained release effect and conformed to the Maloid-Banakar release model (R2=0.9967). The results showed that FA-CS-QUE-LP had higher inhibition rates on HepG2 cells than QUE-Sol (P < 0.01). There was a significant difference in AUC, t1/2, CL and other pharmacokinetic parameters among QUE-LP, FA-CS-QUE-LP, and QUE-Sol (P < 0.05). In in vivoantitumor activity study, the weight inhibition rate and volume inhibition rate of FA-CS-QUE-LP were 30.26% and 37.35%, respectively.Conclusion: FA-CS-QUE-LP exhibited a significant inhibitory effect on HepG2 cells, influenced the pharmacokinetics of quercetin in mice, and demonstrated a certain inhibitory effect on S180 tumor-bearing mice, thus offering novel avenues for cancer treatment.

A record net gain of 22-dB and improved twice gain-per-unit length (GPUL) of 110 dB/m in a 13-μm core, 20-cm length, and 300-nm broadband single-mode Cr-doped crystalline core fibers (CCFs) are demonstrated by AI-assisted image recognition growth. The gain is significantly higher compared to the previous core of 25-μm. Both high gain and GPUL are achieved thanks to a novel AI-assisted, image recognition applied to the precise control of the molten zone with luminance of 2.39 cd/m2 for fabricating smaller core diameter of 13-μm. The AI-assisted growth is performed by capturing images of the molten zone, RGB image recognition, luminance calculation, and real-time luminance adjustment within required value. The gain of several core diameters and fiber lengths has been thereafter simulated, which agrees with measured result. The smaller core of 13-μm exhibits higher coupling efficiency that is 65% the previous core of 25-μm. A lower noise figure of 4.2-dB is also obtained. In this study, the proposed CDFs looks promising for broadband fiber amplifiers.

Histopathological images provide the medical evidences to help the disease diagnosis. However, pathologists are not always available or are overloaded by work. Moreover, the variations of pathological images with respect to different organs, cell sizes and magnification factors lead to the difficulty of developing a general method to solve the histopathological image classification problems. To address these issues, we propose a novel cross-scale fusion (CSF) transformer which consists of the multiple field-of-view patch embedding module, the transformer encoders and the cross-fusion modules. Based on the proposed modules, the CSF transformer can effectively integrate patch embeddings of different field-of-views to learn cross-scale contextual correlations, which represent tissues and cells of different sizes and magnification factors, with less memory usage and computation compared with the state-of-the-art transformers. To verify the generalization ability of the CSF transformer, experiments are performed on four public datasets of different organs and magnification factors. The CSF transformer outperforms the state-of-the-art task specific methods, convolutional neural network-based methods and transformer-based methods.

As a dual three-phase permanent magnet synchronous motor (DTP-PMSM) is fed by two DC voltage source inverters (VSIs), the DC-link voltages may become unbalanced. The conventional direct torque control (DTC) schemes under unbalanced DC-link voltages generate large current harmonics and unbalanced stator currents in two sets of windings, which reduce the efficiency and maximum output power of DTP-PMSM. To solve these problems, DTC schemes considering unbalanced DC-link voltages are proposed in this study to improve the performance of DTP-PMSM. First, the variation law of voltage vectors decomposed into the $\bm{\alpha} \text{-} \bm{\beta}$ and x-y subspaces is analyzed under unbalanced DC-link voltages. Then, two or three sets of new virtual vectors synthesized by three nonzero vectors are developed to retain the voltage in the x-y subspace at zero. Moreover, to extend the applicable range, two DTC schemes switch the virtual vector sets with changes in the DC-link voltage ratio (udc1/udc2). Switching points are theoretically analyzed. Subsequently, their performances are analyzed, including DC-link voltage utilization and ripple current. Finally, experiments are performed to verify the feasibility of the proposed schemes in steady-state and dynamic conditions.

The concept of body constitution (BC) is a core notion in traditional Chinese medicine, used in diagnosis, treatment, and prevention; however, there is little standardization in terms of definitions and measurements. To improve standardization, constructive questionnaires have been developed to classify the various BC types. One of the most commonly used is the Constitution in Chinese Medicine Questionnaire (CCMQ). Despite including nine BCs, CCMQ lacks the blood-vacuity constitution, although it is often noted in clinical practice. In this study, we have modified the original CCMQ to include the blood-vacuity constitution and amended the language to better suit the Taiwan population. The revised questionnaire was given to a panel of experts to check for content validity, and then distributed to volunteers for completion. The reliability analysis, based on 512 valid questionnaires, achieved a Cronbach’s alpha value of 0.65–0.86. The content validity index scores ranged from 60 to 100. In addition, we collected demographic data from our volunteers and found that BMI, gender, exercise frequency, disease status, allergies, and psychiatric disorder status may impact the body constitution. Collectively, our study presents an expanded version of the CCMQ which includes the blood-vacuity constitution, and has been validated among the Taiwanese population. Demographic data also demonstrates possible relationships among BC, lifestyle and diseases.

A novel power boundary analysis-based black start control (PBBC) strategy of PMSG-based wind turbines (WTs) is proposed in this paper to reduce the outage time of WTs and improve the reliability of wind power supply. The active power-rotor speed maximum boundary is quantified to estimate the kinetic energy (KE) reserve limit and the power response capability of the WTs during the black start period. In the power boundary analysis, the power response capability will collapse as the rotor speed increases to maximum. To achieve the trade-off between high KE reserve and power response capability, an overloaded power response control method during the black start period is developed, which can fully utilize the reserved KE. Moreover, the dc-bus capacitor charging model of WTs with initial KE is constructed, which can accurately represent the required KE of the WTs black start. The WTs are coordinated to achieve black start with the minimum backup power supply for the safety devices and auxiliary load according to the capacitor charging model. The proposed PBBC strategy can maximize the KE reserve of the WT and improve the power response capability, as well as reduce the backup power cost during the black start period.

Polyurethane (PU) foams, pivotal in modern life, face challenges suh as fire hazards and environmental waste burdens. The current reliance of PU on potentially ecotoxic halogen-/phosphorus-based flame retardants impedes large-scale material recycling. Here, our demonstrated controllable catalytic cracking strategy, using cesium salts, enables self-evolving recycling of flame-retardant PU. The incorporation of cesium citrates facilitates efficient urethane bond cleavage at low temperatures (160 °C), promoting effective recycling, while encouraging pyrolytic rearrangement of isocyanates into char at high temperatures (300 °C) for enhanced PU fire safety. Even in the absence of halogen/phosphorus components, this foam exhibits a substantial increase in ignition time (+258.8%) and a significant reduction in total smoke release (−79%). This flame-retardant foam can be easily recycled into high-quality polyol under mild conditions, 60 °C lower than that for the pure foam. Notably, the trace amounts of cesium gathered in recycled polyols stimulate the regenerated PU to undergo self-evolution, improving both flame-retardancy and mechanical properties. Our controllable catalytic cracking strategy paves the way for the self-evolutionary recycling of high-performance firefighting materials.

Natural disasters such as hurricane would cause damages in distribution systems (DSs), resulting in large-scale blackouts. In order to reduce socioeconomic losses under blackouts, it is critical to enhance resilience of DSs through a series of restoration means. Microgrids (MGs) are considered as viable and critical sources to provide emergency power support, and can facilitate restoration of DSs when properly coordinated with the utility-owned control measures. Instead of direct coordination control, this paper proposes a transactive energy (TE) framework for load restoration of DSs, under which the distribution system operator (DSO) organizes a TE market with MGs to trade emergency energy through price signals. Moreover, in order to consider renewable generation uncertainties, a distributionally robust optimization (DRO)-based sequential load restoration model is formulated under the TE framework to maximize the expected restored load with respect to the worst-case probability distribution. Case studies were conducted on the modified IEEE 123 node system. Simulation results show that the proposed strategy can effectively coordinate MGs scheduling to assist in load restoration. Moreover, as compared to the conventional stochastic programming model, the proposed DRO model is less sensitive to the number of scenarios modeled and can produce a more robust solution against out-of-sample scenarios.

Semantic segmentation of basal cell carcinoma (BCC) from full-field optical coherence tomography (FF-OCT) images of human skin has received considerable attention in medical imaging. However, it is challenging for dermatopathologists to annotate the training data due to OCT’s lack of color specificity. Very often, they are uncertain about the correctness of the annotations they made. In practice, annotations fraught with uncertainty profoundly impact the effectiveness of model training and hence the performance of BCC segmentation. To address this issue, we propose an approach to model training with uncertain annotations. The proposed approach includes a data selection strategy to mitigate the uncertainty of training data, a class expansion to consider sebaceous gland and hair follicle as additional classes to enhance the performance of BCC segmentation, and a self-supervised pre-training procedure to improve the initial weights of the segmentation model parameters. Furthermore, we develop three post-processing techniques to reduce the impact of speckle noise and image discontinuities on BCC segmentation. The mean Dice score of BCC of our model reaches 0.503±0.003, which, to the best of our knowledge, is the best performance to date for semantic segmentation of BCC from FF-OCT images.

Conventional model-free predictive current control (MFPCC) using an extended state observer (ESO) based on ultra-local model significantly decreases the sensitivity to partial motor parameters. However, the performance and stability of ESO will be degraded in the face of amplified high-frequency (HF) perturbation caused by inductance parameter mismatch. To solve the above problem, an enhanced ESO-based MFPCC method is proposed in this article. On the basis of constructing the discrete equation of ESO, an accurate and fast calculation of inductance parameter is proposed, which is real-time convergence to the actual value by deadbeat control principle, and fed back to the lumped disturbance and current observation calculation. It effectively eliminates the effects of HF perturbation on ESO. Simultaneously, the inductance and compensated lumped disturbance obtained from ESO are provided to the MFPCC command voltage calculation module, further improving the parameter robustness of MFPCC. The stability of the proposed ESO is ensured by the design of observer gain. Experiments are carried out on 1-kW surfaced permanent magnet synchronous machines test bench. Compared with the conventional ESO-based MFPCC method, the proposed method is superior in dynamic response, steady-state performance, and parameter robustness.

Data augmentation is a popular way for few-shot learning (FSL). It generates more samples as supplements and then transforms the FSL task into a common supervised learning problem for a solution. However, most data-augmentation-based FSL approaches only consider the prior visual knowledge for feature generation, thereby leading to low diversity and poor quality of generated data. In this study, we attempt to address this issue by incorporating both prior visual and prior semantic knowledge to condition the feature generation process. Inspired by some genetic characteristics of semi-identical twins, a novel multimodal generative FSL approach was developed named semi-identical twins variational autoencoder (STVAE) to better exploit the complementarity of these modality information by considering the multimodal conditional feature generation process as a process that semi-identical twins are born and collaborate to simulate their father. STVAE conducts feature synthesis by pairing two conditional variational autoencoders (CVAEs) with the same seed but different modality conditions. Subsequently, the generated features of two CVAEs are considered as semi-identical twins and adaptively combined to yield the final feature, which is considered as their fake father. STVAE requires that the final feature can be converted back into its paired conditions while ensuring these conditions remain consistent with the original in both representation and function. Moreover, STVAE is able to work in the partial modality-absence case due to the adaptive linear feature combination strategy. STVAE essentially provides a novel idea to exploit the complementarity of different modality prior information inspired by genetics in FSL. Extensive experimental results demonstrate that our work achieves promising performances in comparison to the recent state-of-the-art approaches, as well as validate its effectiveness on FSL under various modality settings.

Downscaling nighttime light (NTL) from satellite imagery presents valuable applications at a more detailed spatial scale, especially in the realms of urban expansion and socio-economic assessment. Nevertheless, due to the complexity of geographical conditions and uncertainties in the relationships among multiple factors, the precision of NTL downscaling often encounters constraints. In this work, an incorporated multifactor geographically neural network weighted regression (MF-GNNWR) NTL downscaling framework is proposed to solve the spatial nonstationarity in high-heterogeneous urban areas, which mainly uses geographically neural network weighted regression (GNNWR) combined with multiple factors including surface physical characteristics, socio-economic attributes, and human activities to improve the accuracy of NTL, particularly in urban regions with complicated land cover. The findings illustrate that the MF-GNNWR framework displays finer downscaling accuracy on different land cover, effectively enhancing data quality. Notably, our findings underscore the pronounced influence of socio-economic and human activity factors on NTL downscaling. Comparative analysis against several alternative downscaling methodologies reveals that the MF-GNNWR framework outperforms them, exhibiting a remarkable 23.10% improvement in the Pearson correlation coefficient (r) and achieving a root-mean-square error (RMSE) of $16.95~\text {{nW/c}{m}}^{2}{/\text {sr}}$ , and after residual compensation, r continue s to increase by 1.5%, while RMSE decreases by $0.157~\text {nW/cm}^{2}{/\text {sr}}$ . These findings highlight the efficacy of the proposed framework in downscaling NTL, underscoring its advantages and practical utility.

The fault of the collector cable in a wind farm (WF) may lead to the tripping of wind turbines (WTs), and thus threaten the secure and stable operation of the WF. In this article, a topology reconfiguration-based optimal fault-recovery control (TROFC) scheme is proposed for the WF collection system to isolate the fault while keeping WTs within WF uninterrupted operation. The proposed TROFC scheme also involves two-stage control which includes the offline topology optimization and online fault recovery. In the first stage, an integer second-order cone programming-based optimization method is used to determine the location and number of contact cables offline, which takes the contact cable cost, average voltage, and current deviation of WF system into consideration. In the second stage, the closed-loop current and limit closing time are quantified. The online control scheme is designed based on quantified closed-loop current and limit closing time to enhance the safe operation capacity of a WF during the topology reconfiguration period. The order of topology reconfiguration is closed-loop operation followed by de-loop operation. The effectiveness of the proposed TROFC scheme is validated in MATLAB/Simulink.

The spatial-resolved capacity of a phase-sensitive optical time-domain reflectometry ( $\varphi $ -OTDR) is limited by the optical pulse width, which increases the localization errors when measuring vibrations with range shorter than that determined by the pulse width. To overcome this problem, a high spatial-resolved $\varphi $ -OTDR with a differential pulse width sequence is proposed. The centimeter spatial resolution sensing can be realized by using meter equivalent pulses. This scheme utilizes the time-domain waveform subtraction obtained from the same pulse repetition rate and different pulse widths to form the differential vibration information at each fiber location. First, we provide a comprehensive treatment to prove that the proposed scheme can improve the spatial-resolved capacity of the $\varphi $ -OTDR system. The spatial resolution of 1.08m for 60/50ns pulse sequence is realized with a 30cm vibration source. Second, in a further proof-of-concept experiment, the spatial resolution is achieved up to 0.55m with a 55/50ns pulse sequence. It is demonstrated that the ability of spatial-resolved is enhanced greatly without any hardware modifications, compared to that required in the conventional method. This method holds great significance in accurately locating vibration signals within the narrow range of the distributed vibration sensor.

Speech-driven facial animation is a challenging problem where each input audio can have multiple plausible facial outputs, leading to overly smooth results. Although the two-stage framework of audio-to-motion model and neural rendering models can partially mitigate this issue, it lacks crucial details like emotions and wrinkles. To overcome these limitations, we introduce a variational motion generator with implicit memory. By incorporating implicit memory into the audio-to-motion model, we capture high-level semantics in the shared latent space of audio expressions, resulting in accurate and expressive facial landmark generation. Next, we introduce attention with time bias to effectively maintain the consistency of audio motion and adopt a periodic position encoding strategy to provide summarization capability for longer audio sequences. Experimental results demonstrate that our approach outperforms previous methods, yielding more extensive and realistic speech-driven facial animation.

Vehicle identification number (VIN) is a unique serial number used to identify individual vehicles across various applications. The first crucial step in automatically collecting VINs is to accurately localize the VIN area. In this article, we present a novel VIN detection approach called Mirrored EAST (MEAST) based on an efficient and accurate scene text (EAST) Detection framework. MEAST learns to exploit the spatial consistency between an image and its mirrored version to improve localization performance, and employs a lighter but more discriminative backbone network to improve its applicability in mobile scenarios. To evaluate the VIN detection performance, we constructed a large-scale VIN image dataset named CQU-VD20 K, consisting of 20 000 VIN images in real scenarios. Based on this dataset, we have conducted a comprehensive empirical study of VIN detection. The results demonstrate the superiority of MEAST over other methods in VIN detection. Additionally, we also conducted extended experiments on a license plate dataset named CCPD-Rotate, which confirms the effectiveness of our approach in other industrial inspection tasks.

The knowledge about a network model is essential for control. However, the topology of wind farms (WFs) may be changed due to random off-grid events of wind turbines. This article develops a two-stage local method for WFs to achieve the Var/Volt coordinated optimization with consideration of topology changes, which provides an efficient solution when limited data are available for identification and control. In the first stage, considering the operational characteristics of WFs, a mixed integer optimization problem is formulated in each local controller to uncover the physical network topology. The problem is then solved and driven toward the optimum by an online recursive search algorithm. The proposed algorithm is intended to identify the physical network layout only using local measurements at no loss of accuracy. In the second stage, a completely decentralized Var/Volt control method is designed with the updated topological knowledge, aiming to achieve a fast Var response and to reduce voltage deviation caused by the stochastic wind power and voltage disturbance of external girds. Case studies in MATLAB/Simulink verify the robustness and effectiveness of the proposed method within a wide range of working conditions.

In industrial production, the speed of the motor is constantly changing due to production demands. However, traditional diagnosis methods cannot guarantee their performance under all circumstance, especially the sharp changing speed condition. To tackle this issue, this article proposes an adaptive method that can be used for multifaults diagnosis of motor under variety working condition, especially for the dramatic changing and unstable working states. First, a time–frequency (TF) parameter and resolution adaptive algorithm is proposed for signal preprocessing. Second, the well-processed signal is resampled by TF diagram and peak search method to eliminate the effect of sudden change in speed. A proposed perception matching algorithm based on the symmetrized dot pattern and convolutional neural networks is implemented for improving the accuracy. Finally, an adaptive model with unified diagnose process is developed to improve the accuracy, efficiency, and practicality of the diagnose model. Compared with other state-of-art methods, the results show the out-performance of the proposed method under both in steady and transient state, especially in the case that the speed is changing fast.

Photomemristors have been regarded as one of the most promising candidates for next-generation hardware-based neuromorphic computing due to their potentials of fast data transmission and low power consumption. However, intriguingly, so far, photomemristors seldom display truly nonvolatile memory characteristics with high light sensitivity. Herein, we demonstrate ultrasensitive photomemristors utilizing two-dimensional (2D) Ruddlesden–Popper (RP) perovskites with a highly polar donor–acceptor-type push–pull organic cation, 4-(5-(2-aminoethyl)thiophen-2-yl)benzonitrile+(EATPCN+), as charge-trapping layers. High linearity and almost zero-decay retention are observed in (EATPCN)2PbI4devices, which are very distinct from that of the traditional 2D RP perovskite devices consisting of nonpolar organic cations, such as phenethylamine+(PEA+) and octylamine+(OA+), and traditional 3D perovskite devices consisting of methylamine+(MA+). The 2-fold advantages, including desirable spatial crystal arrangement and engineered energetic band alignment, clarify the mechanism of superior performance in (EATPCN)2PbI4devices. The optimized (EATPCN)2PbI4photomemristor also shows a memory window of 87.9 V and an on/off ratio of 106with a retention time of at least 2.4 × 105s and remains unchanged after >105writing–reading–erasing–reading endurance cycles. Very low energy consumptions of 1.12 and 6 fJ for both light stimulation and the reading process of each status update are also demonstrated. The extremely low power consumption and high photoresponsivity were simultaneously achieved. The high photosensitivity surpasses that of a state-of-the-art commercial pulse energy meter by several orders of magnitude. With their outstanding linearity and retention, rabbit images have been rebuilt by (EATPCN)2PbI4photomemristors, which truthfully render the image without fading over time. Finally, by utilizing the powerful ∼8 bits of nonvolatile potentiation and depression levels of (EATPCN)2PbI4photomemristors, the accuracies of the recognition tasks of CIFAR-10 image classification and MNIST handwritten digit classification have reached 89% and 94.8%, respectively. This study represents the first report of utilizing a functional donor–acceptor type of organic cation in 2D RP perovskites for high-performance photomemristors with characteristics that are not found in current halide perovskites.

Alkoxy side chain engineering on the β-position of the thienothiophene units of Y6 derivatives plays a vital role in improving photovoltaic performances with simultaneously increasing open-circuit voltage (Voc) and fill factor (FF). In this work, we prepared a series of asymmetric non-fullerene acceptors (NFAs) by introducing alkoxy side chains and phenoxy groups on the state-of-the-art Y6-derivative BTP-BO-4F. For the comparison, 2O-BO-4F with a symmetric alkoxy side chain on the outer thiophene units and BTP-PBO-4F with an asymmetric N-attached phenoxy alkyl chain on the pyrrole ring are synthesized from BTP-BO-4F. Thereafter, we construct four asymmetric NFAs by introducing different lengths of linear/branched alkoxy chains on the β-position of the thienothiophene units of BTP-PBO-4F. The resulting NFAs, named L10-PBO, L12-PBO, B12-PBO, and B16-PBO (L = linear and B = branched alkoxy side chains), are collectively called OR-PBO-series. Unexpectedly, all OR-PBO NFAs exhibit strong edge-on molecular packing and weaker π–π interactions in the film state, which diminish the charge transfer in organic solar cell (OSC) devices. As a consequence, the optimal devices of OR-PBO-based binary blends show poor photovoltaic performances [power conversion efficiency (PCE) = 6.52–9.62%] in comparison with 2O-BO-4F (PCE = 12.42%) and BTP-PBO-4F (PCE = 15.30%) reference blends. Nevertheless, the OR-PBO-based binary devices show a higher Vocand smaller Vloss. Especially, B12-PBO- and B16-PBO-based devices achieve Vocover 1.00 V, which is the highest value of Y-series OSC devices to the best of our knowledge. Therefore, by utilizing higher Vocof OR-PBO binary blends, B12-PBO and B16-PBO are incorporated into the PM6:BTP-PBO-4F-based binary blend and fabricated ternary devices. As a result, the PM6:BTP-PBO-4F:B12-PBO ternary device delivers the best PCE of 15.60% with an increasing Vocand FF concurrently.

Because imbalanced power will cause the loss of the propulsion motor of the railway vehicle, and the increase in temperature will shorten the service life of the electric vehicle. Not only this but also increase the cost of electricity and maintenance. In the past, the industry only focused on methods to improve power quality such as load capacity, relay setting, and harmonic resolution. Now, the consider of three-phase unbalance rate (TPUR) must be applied. I propose special transformers wiring (STW) to improve the three unbalance rates and provide different transformer wiring methods. According to the IEEE Committee, in the future, power companies will need to install balanced relay stations to improve three-phase unbalance rate. The internal regulations of Taipower must be less than 4.5% (voltage unbalance rate (NPSUR) of 2.5% and motor temperature rise of 12.5%).The derivation of the transformer “three-phase unbalance rate” model is the focus of the railway system. This research is based on the model derivation of different wiring methods to improve the problem caused by the three-phase imbalance and improve the using life of the train. They pointed out that Scott, Le-Blanc, and Modified-Woodbridge, the three wiring methods can be applied to future railway system and reduce the three-phase unbalance rate. They are in line with the IEEE standard ( %PUV) of less than 2%.

Sulfur dioxide is a colorless gas with strong pungent odor, and prolonged inhalation can damage the respiratory system and heart. Precise measurement of sulfur dioxide is very important for the human health. While the current sulfur dioxide sensor works at a high temperature and is easily interfered by humidity. Therefore, in this work, we used a solution method to coat perovskite nanocrystals Cs3Cu2I5 with metal oxides Fe2O3 and obtained the room-temperature sulfur dioxide gas-sensitive materials Cs3Cu2I5@Fe2O3 nanocrystals. Meanwhile, combined with machine learning, the sensor has the ability of self-calibrating under different humidity environment, realizing the high precision and anti-interference measurement of sulfur dioxide. In addition, in this structure, the coating Fe2O3 not only helps to improve water stability but also transfers the interaction of sulfur dioxide to the inner Cs3Cu2I5 and lowers the gas-sensitive temperature to room temperature. The response/recovery time of the sulfur dioxide sensor based on Cs3Cu2I5@Fe2O3 is 31/816 s, and the sensitivity is 0.19 at 10 ppm. Then, intelligent classification algorithm was used for recognition, and the accurate recognition rate was up to 95.0%. Furthermore, density functional theory (DFT) was implemented to reveal the gas-sensitive mechanism that sulfur dioxide was adsorbed by Cs3Cu2I5@Fe2O3 nanocrystals, the good response was attributed to the band structure changes significantly, and the hybridization of electron orbitals was appeared between gas and nanocrystals. We believe that the sensor will have potential in sulfur dioxide, and the idea of using machine learning to intelligent eliminate humidity interference can also be extended to other gas sensor.

Various additives have been introduced to assist in film preparation and defect passivation. Herein, fluoroiodobenzene (FIB) molecules with different numbers of F atoms were incorporated into perovskite films to optimize the film quality as well as passivate defects. Based on the calculation and experimental results, it was found that the FIB additives were inclined to exist at the bottom of the film because of the strong affinity between F atoms stemming from FIB molecules and O atoms stemming from TiO2, especially for molecules with more F atoms. By optimization of the FIB molecule, the perovskite film crystallinity was significantly improved, the carrier lifetimes were prolonged, and the charge extraction ability was also enhanced. The device with FIB with one F atom achieved a photoelectrical conversion efficiency as high as 22.89% with a Vocof 1.118 V, fill factor (FF) of 80.44%, and Jscof 25.45 mA cm–2, which was much higher than that of the control device with an efficiency of 20.87%. Furthermore, FIB molecules with three and five F atoms also achieved higher efficiency than that of the control device. The devices with FIB molecules showed better stability than the devices without additives. The unencapsulated devices with FIB additives held 90% of their original efficiencies in an ambient environment with a temperature of 15–25 °C and a relative humidity of 20–30%, while the control device dropped to 76% after more than 1000 h.

• A novel sustainable bio-based flame-retardant system for biomass RPUF was synthesized. • Efficient flame retardancy and smoke suppression were achieved for the composite. • Thermal insulation performance was kept in good condition after being treated in high-humidity environment with 85% R.H. for a week. Bio-based polyurethane foam has attracted increasing attentions due to eco-friendliness and fossil feedstock issues. However, the inherent flammability limits its application in different fields. Herein, we demonstrate a green bio-based flame-retardant system to fabricate polyurethane foam composite with durable flame retardancy, smoke suppression, and thermal insulation property. In this system, the green bio-based polyol (VED) with good reactivity and compatibility plays a role of flame retardant and EG acts as a synergistic filler. As a result, the LOI value of foam composite increased to 30.5 vol.% and it achieved a V-0 rating in the UL-94 vertical burning test. Additionally, the peak heat release rate (pHRR) and the total smoke production (TSP) decreased by 66.1% and 63.4%, respectively. Furthermore, the foam composite maintained durable flame retardancy after accelerated thermal aging test, whose thermal-insulating property was maintained even after being treated in high-humidity environment with 85% R.H. for a week. This work provides a facile strategy for durable flame retardancy and long-term thermal insulation performance, and creates opportunities for the practical applications of bio-based foam composites. [Display omitted] [ABSTRACT FROM AUTHOR]

Scoring functions for protein–ligand interactions play a critical role in structure-based drug design. Owing to the good balance between general applicability and computational efficiency, knowledge-based scoring functions have obtained significant advancements and achieved many successes. Nevertheless, knowledge-based scoring functions face a challenge in utilizing the experimental affinity data and thus may not perform well in binding affinity prediction. Addressing the challenge, we have proposed an improved version of the iterative knowledge-based scoring function ITScore by considering binding affinity information, which is referred to as ITScoreAff, based on a large training set of 6216 protein–ligand complexes with both structures and affinity data. ITScoreAff was extensively evaluated and compared with ITScore, 33 traditional, and 6 machine learning scoring functions in terms of docking power, ranking power, and screening power on the independent CASF-2016 benchmark. It was shown that ITScoreAff obtained an overall better performance than the other 40 scoring functions and gave an average success rate of 85.3% in docking power, a correlation coefficient of 0.723 in scoring power, and an average rank correlation coefficient of 0.668 in ranking power. In addition, ITScoreAff also achieved the overall best screening power when the top 10% of the ranked database were considered. These results demonstrated the robustness of ITScoreAff and its improvement over existing scoring functions.

Camptothecin (CPT) and its derivatives have attracted worldwide attention because of their notable anticancer activity. However, the growing demand for CPT in the global pharmaceutical industry has caused a severe shortage of CPT-producing plant resources. In this study, phytochemical analysis of Nothapodytes tomentosa results in the isolation and identification of CPT (13) and 16 analogues (1 – 12, 14 – 17), including a new (1) and five known (9, 10, 12, 15 , and 17) CPT analogues with an open E-ring. In view of the potential anticancer activity of CPT analogues with an open E-ring, the fragmentation pathways and mass spectra profiles of these six CPT analogues (1, 9, 10, 12, 15 , and 17) are investigated, providing a reference for the rapid detection of these compounds in other plants. Furthermore, based on the fragmentation patterns of CPT (13) and known analogues (2 – 8, 11, 14, 16, 18 – 26), the distribution and content of these compounds in different tissues of N. tomentosa, N. nimmoniana, Camptotheca acuminata , and Ophiorrhiza japonica are further studied. Our findings not only provide an alternative plant resource for further expanding the development and utilization of CPT and its analogues, but also lay a foundation for improving the utilization of known CPT-producing plant resources. [ABSTRACT FROM AUTHOR]

Facial image-based kinship verification is a rapidly growing field in computer vision and biometrics. The key to determining whether a pair of facial images has a kin relation is to train a model that can enlarge the margin between the faces that have no kin relation while reducing the distance between faces that have a kin relation. Most existing approaches primarily exploit duplet (i.e., two input samples without cross pair) or triplet (i.e., single negative pair for each positive pair with low-order cross pair) information, omitting discriminative features from multiple negative pairs. These approaches suffer from weak generalizability, resulting in unsatisfactory performance. Inspired by human visual systems that incorporate both low-order and high-order cross-pair information from local and global perspectives, we propose to leverage high-order cross-pair features and develop a novel end-to-end deep learning model called the adaptively weighted $k$ -tuple metric network (AW $k$ -TMN). Our main contributions are three-fold. First, a novel cross-pair metric learning loss based on $k$ -tuplet loss is introduced. It naturally captures both the low-order and high-order discriminative features from multiple negative pairs. Second, an adaptively weighted scheme is formulated to better highlight hard negative examples among multiple negative pairs, leading to enhanced performance. Third, the model utilizes multiple levels of convolutional features and jointly optimizes feature and metric learning to further exploit the low-order and high-order representational power. Extensive experimental results on three popular kinship verification datasets demonstrate the effectiveness of our proposed AW $k$ -TMN approach compared with several state-of-the-art approaches. The source codes and models are released.1

We propose a novel approach for achieving joint detections of frequency and direction of arrival (DoA) in wideband using a single sensor based on an active metasurface with programmable transmission states of pass and stop. By integrating two p-i-n diodes with the opposite directions into the proposed single-layer and ultrathin meta-atom, the transmission performance with 10 dB difference between the pass and stop states is realized in the bandwidth from 5.9 to 8.8 GHz using field-circuit cosimulations. Accordingly, random receiving patterns are generated by controlling the programmable metasurface composed of switchable meta-atoms. Afterward, the frequency and direction information of sources located in the far field is detected using the modified algorithm of estimating signal parameters via rotational invariance techniques (ESPRITs) and the compressive sensing method, respectively. A sample of the programmable metasurface is fabricated, and the voltage control system is built up correspondingly. To verify the proposed method’s validity, we conduct three typical categories of experiments: a single source, double sources with different frequencies, and double sources with the same frequency. In all cases, the source information of frequency and direction has been detected previously in measurements within the frequency band from 6.2 to 8.8 GHz.