Search Results (3,863)

Buildings are a significant component of urban space and are essential to smart cities, catastrophe monitoring, and land use planning. However, precisely extracting building polygons from remote sensing images remains difficult because of the variety of building designs and intricate backgrounds. This paper proposes an end-to-end polygon dynamic adjustment algorithm (PDAA) to improve the accuracy and geometric consistency of building contour extraction by dynamically generating and optimizing polygon vertices. The method first locates building instances through the region of interest (RoI) to generate initial polygons, and then uses four core modules for collaborative optimization: (1) the feature enhancement module captures local detail features to improve the robustness of vertex positioning; (2) the contour vertex tuning module fine-tunes vertex coordinates through displacement prediction to enhance geometric accuracy; (3) the learnable redundant vertex removal module screens key vertices based on a classification mechanism to eliminate redundancy; and (4) the missing vertex completion module iteratively restores missed vertices to ensure the integrity of complex contours. PDAA dynamically adjusts the number of vertices to adapt to the geometric characteristics of different buildings, while simplifying the prediction process and reducing computational complexity. Experiments on public datasets such as WHU, Vaihingen, and Inria show that PDAA significantly outperforms existing methods in terms of average precision (AP) and polygon similarity (PolySim). It is at least 2% higher than existing methods in terms of average precision (AP), and the generated polygonal contours are closer to the real building geometry. Values of 75.4% AP and 84.9% PolySim were achieved on the WHU dataset, effectively solving the problems of redundant vertices and contour smoothing, and providing high-precision building vector data support for scenarios such as smart cities and emergency response. Full article

The early detection of liver cirrhosis (LC) is crucial due to its high morbidity and mortality in advanced stages. Reliable, non-invasive diagnostic tools are essential for timely intervention. Exhaled human breath, reflecting metabolic changes, offers significant potential for disease diagnosis. This paper focuses on the emerging role of sensor array-based volatile organic compounds (VOCs) analysis of exhaled breath, particularly using electronic nose (e-nose) technology to differentiate LC patients from healthy controls (HCs). This study included 55 participants: 27 LC patients and 28 HCs. Sensor’s measurement data were analyzed using machine learning techniques, such as principal component analysis (PCA), discriminant function analysis (DFA), and support vector machines (SVMs) that were utilized to uncover meaningful patterns and facilitate accurate classification of sensor-derived information. The diagnostic accuracy was thoroughly assessed through receiver operating characteristic (ROC) curve analysis, with specific emphasis on assessing sensitivity and specificity metrics. The e-nose effectively distinguished LC from HC, with PCA explaining 92.50% variance and SVMs achieving 100% classification accuracy. This study demonstrates the significant potential of e-nose technology towards VOCs analysis in exhaled breath, as a valuable tool for LC diagnosis. It also explores feature extraction methods and suitable algorithms for effectively distinguishing between LC patients and controls. This research provides a foundation for advancing breath-based diagnostic technologies for early detection and monitoring of liver cirrhosis. Full article

Background/Objectives: Postoperative deep vein thrombosis (DVT) is a common and serious complication after tibial fracture surgery. This study aimed to develop and evaluate machine learning (ML) models to predict the occurrence of DVT following tibia fracture surgery. Methods: A retrospective analysis was conducted on patients who had undergone surgery for isolated tibial fractures. A total of 42 predictive models were developed using combinations of six ML algorithms—logistic regression, support vector machine, random forest, extreme gradient boosting, Light Gradient Boosting Machine (LightGBM), and neural networks—and seven feature selection methods, including SHapley Additive exPlanations (SHAP), Least Absolute Shrinkage and Selection Operator (LASSO), Boruta, recursive feature elimination, univariate filtering, and full-variable inclusion. Model performance was assessed based on discrimination, quantified by the area under the receiver operating characteristic curve (AUC-ROC), and calibration, measured using Brier scores, with internal validation performed via bootstrapping. Results: Of 471 patients, 80 (17.0%) developed postoperative DVT. The ML models achieved high overall accuracy in predicting DVT. Twenty-four models showed similarly excellent discrimination (pairwise AUC comparisons, p > 0.05). The top-performing model (random forest with RFE) attained an AUC of ~0.99, while several others (including LightGBM and SVM-based models) also reached AUC values in the 0.97–0.99 range. Notably, support vector machine models paired with Boruta or LASSO feature selection demonstrated the best calibration (lowest Brier scores), indicating reliable risk estimation. The final selected SVM models achieved high specificity (≥95%) with moderate sensitivity (~75–80%) for DVT detection. Conclusions: ML models demonstrated high accuracy in predicting postoperative DVT following tibial fracture surgery. Support vector machine-based models showed particularly favorable discrimination and calibration. These results suggest the potential utility of ML-based risk stratification to guide individualized prophylaxis, warranting further validation in prospective clinical settings. Full article

Landslide susceptibility assessment (LSA) plays a crucial role in disaster prevention and mitigation. Traditional random selection of non-landslide samples (labeled as 0) suffers from poor representativeness and high randomness, which may include potential landslide areas and affect the accuracy of LSA. To address this issue, this study proposes a novel Landslide Susceptibility Index–based Semi-supervised Fuzzy C-Means (LSI-SFCM) sampling strategy combining membership degrees. It utilizes landslide and unlabeled samples to map landslide membership degree via Semi-supervised Fuzzy C-Means (SFCM). Non-landslide samples are selected from low-membership regions and assigned membership values as labels. This study developed three models for LSA—Convolutional Neural Network (CNN), U-Net, and Support Vector Machine (SVM), and compared three negative sample sampling strategies: Random Sampling (RS), SFCM (samples labeled 0), and LSI-SFCM. The results demonstrate that the LSI-SFCM effectively enhances the representativeness and diversity of negative samples, improving the predictive performance and classification reliability. Deep learning models using LSI-SFCM performed with superior predictive capability. The CNN model achieved an area under the receiver operating characteristic curve (AUC) of 95.52% and a prediction rate curve value of 0.859. Furthermore, compared with the traditional unsupervised fuzzy C-means (FCM) clustering, SFCM produced a more reasonable distribution of landslide membership degrees, better reflecting the distinction between landslides and non-landslides. This approach enhances the reliability of LSA and provides a scientific basis for disaster prevention and mitigation authorities. Full article

Adeno-associated virus (AAV) vectors face a critical translational challenge in ocular gene therapy due to pre-existing neutralizing antibodies (NAbs) whose seroprevalence limits patient eligibility. Standard NAb detection using non-ocular cell models (Human Embryonic Kidney 293T) may inadequately predict retinal transduction inhibition due to cell type-related variations in receptor usage and immunogenicity. This study established parallel NAb detection platforms utilizing human retinal pigment epithelial (ARPE-19) cells and standard 293T cells to systematically evaluate clinical serum samples against ophthalmologically relevant AAV serotypes (2, 5, 8, 9) via luciferase reporter-based transduction inhibition assays. Comparative analysis demonstrated ARPE-19 exhibited 42–48% higher NAb titers against AAV5/9 compared to 293T cells, with distinct serotype-biased neutralization hierarchies observed between cellular models. Furthermore, female-derived sera exhibited significantly elevated NAbs against particular serotypes in the ARPE-19 system. Critically, inter-serotype cross-neutralization correlation patterns differed substantially between cellular platforms. These findings demonstrate that physiologically relevant retinal cellular models provide essential immunological profiling data, revealing NAb characteristics obscured in standard assays. Consequently, employing retinal cell-based platforms is crucial for optimizing AAV serotype selection, patient stratification, and predicting clinical outcomes in ocular gene therapy. Full article

Marine coral sand–clay mixtures (MCCM) are promising green fill materials in civil engineering projects, where their strength characteristics play a vital role in ensuring structural safety and stability. To investigate these properties, a series of triaxial shear tests were performed under diverse conditions, including variations in asperity spacing, asperity height, the number of reinforcement layers, confining pressure, and axial strain. This experimental campaign yielded a robust strength dataset for MCCM. Utilizing this dataset, several predictive models were developed, including a standard Support Vector Machine (SVM), an SVM optimized via Genetic Algorithm (GA-SVM), an SVM enhanced by Particle Swarm Optimization (PSO-SVM), and a hybrid model incorporating Logical Development Algorithm preprocessing a SVM model (LDA-SVM). Among these models, the LDA-SVM model exhibited the best performance, achieving a test RMSE of 1.67245 and a correlation coefficient (R) of 0.996, demonstrating superior prediction accuracy and strong generalization ability. Sensitivity analyses revealed that asperity spacing, asperity height, and confining pressure are the most influential factors affecting MCCM strength. Moreover, an explicit empirical equation was derived from the LDA-SVM model, allowing practitioners to estimate strength without relying on complex machine learning tools. The results of this study offer practical guidance for the optimized design and safety evaluation of MCCM in civil engineering applications. Full article

In the context of the new developmental philosophy, this study aimed to address the bottleneck of regional sustainable development; it constructs a three-system evaluation indicator system for Industrial Structure Optimization (ISO), Ecological Environment Management (EEM), and Socio-economic Development (SED), based on panel data from 20 cities in the Western Taiwan Straits Economic Zone between 2011 and 2023. To reveal how the synergistic development of the three subsystems in different domains can achieve sustainable development through their interactions and to analyze the dynamic patterns of the three subsystems, this study employed the panel vector autoregression (PVAR) model to examine the interactions between subsystems. Additionally, drawing on the framework of evolutionary economics, the study quantified the temporal evolution and spatial characteristics of the coupling coordination level among the three subsystems based on the results of the degree of coupling coordination model. The results indicate the following: (1) ISO shows a significant upward trend, EEM slightly declines, and SED experiences minor fluctuations before accelerating. (2) ISO, EEM, and SED exhibited self-reinforcing effects. (3) The degree of coupling, coordination, and coupling coordination all exhibit a trend of “fluctuating and increasing initially, followed by steady growth”. The spatial patterns of the degree of coupling, coordination, and coupling coordination have shifted from “decentralized” to “centralized”, with clear signs of synergistic development. (4) The difference in the degree of coupling coordination along the north–south direction remained the primary factor contributing to inter-regional disparities. Regions with the higher degrees of coupling coordination were concentrated in the southeastern coastal areas, while those with the lower degrees of coupling coordination appeared in the northeastern mountainous areas and southwestern coastal areas. (5) The spatial connection in the strength of the degree of coupling coordination has gradually increased, with notable intra-provincial connections and weakened inter-city connections across the province. The study’s results provided decision-making references for the construction of a sustainable development community. Full article

This paper presents some of the most significant findings in the design of a hydraulic servomechanism for flight controls, which were primarily achieved by the first author during his activity in an aviation institute. These results are grouped into four main topics. The first one outlines a classical theory, from the 1950s–1970s, of the analysis of nonlinear automatic systems and namely the issue of absolute stability. The uninformed public may be misled by the adjective “absolute”. This is not a “maximalist” solution of stability but rather highlights in the system of equations a nonlinear function that describes, for the case of hydraulic servomechanisms, the flow-control dependence in the distributor spool. This function is odd, and it is therefore located in quadrants 1 and 3. The decision regarding stability is made within the so-called Lurie problem and is materialized by a matrix inequality, called the Lefschetz condition, which must be satisfied by the parameters of the electrohydraulic servomechanism and also by the components of the control feedback vector. Another approach starts from a classical theorem of V. M. Popov, extended in a stochastic framework by T. Morozan and I. Ursu, which ends with the description of the local and global spool valve flow-control characteristics that ensure stability in the large with respect to bounded perturbations for the mechano-hydraulic servomechanism. We add that a conjecture regarding the more pronounced flexibility of mathematical models in relation to mathematical instruments (theories) was used. Furthermore, the second topic concerns, the importance of the impedance characteristic of the mechano-hydraulic servomechanism in preventing flutter of the flight controls is emphasized. Impedance, also called dynamic stiffness, is defined as the ratio, in a dynamic regime, between the output exerted force (at the actuator rod of the servomechanism) and the displacement induced by this force under the assumption of a blocked input. It is demonstrated in the paper that there are two forms of the impedance function: one that favors the appearance of flutter and another that allows for flutter damping. It is interesting to note that these theoretical considerations were established in the institute’s reports some time before their introduction in the Aviation Regulation AvP.970. However, it was precisely the absence of the impedance criterion in the regulation at the appropriate time that ultimately led, by chance or not, to a disaster: the crash of a prototype due to tailplane flutter. A third topic shows how an important problem in the theory of automatic systems of the 1970s–1980s, namely the robust synthesis of the servomechanism, is formulated, applied and solved in the case of an electrohydraulic servomechanism. In general, the solution of a robust servomechanism problem consists of two distinct components: a servo-compensator, in fact an internal model of the exogenous dynamics, and a stabilizing compensator. These components are adapted in the case of an electrohydraulic servomechanism. In addition to the classical case mentioned above, a synthesis problem of an anti-windup (anti-saturation) compensator is formulated and solved. The fourth topic, and the last one presented in detail, is the synthesis of a fuzzy supervised neurocontrol (FSNC) for the position tracking of an electrohydraulic servomechanism, with experimental validation, in the laboratory, of this control law. The neurocontrol module is designed using a single-layered perceptron architecture. Neurocontrol is in principle optimal, but it is not free from saturation. To this end, in order to counteract saturation, a Mamdani-type fuzzy logic was developed, which takes control when neurocontrol has saturated. It returns to neurocontrol when it returns to normal, respectively, when saturation is eliminated. What distinguishes this FSNC law is its simplicity and efficiency and especially the fact that against quite a few opponents in the field, it still works very well on quite complicated physical systems. Finally, a brief section reviews some recent works by the authors, in which current approaches to hydraulic servomechanisms are presented: the backstepping control synthesis technique, input delay treated with Lyapunov–Krasovskii functionals, and critical stability treated with Lyapunov–Malkin theory. Full article

The objective of this study was to analyze the impact of solar and hybrid dryers on the physicochemical characteristics, fingerprints, and cognitive-sensory perceptions of Mexican consumers of traditional tostadas made with corn of different races. Corn tostadas from different native races were evaluated with solar and hybrid (solar-photovoltaic solar panels) dehydration methods. Proximal chemical quantification, instrumental analysis (color, texture), fingerprint by Fourier transform infrared spectroscopy (FTIR), and sensory-cognitive profile (emotions and memories) and its relationship with the level of pleasure were carried out. The data were evaluated using analysis of variance models, Cochran Q, and an external preference map (PREFMAP). The results showed that the drying method and corn race significantly (p < 0.05) affected only moisture content, lipids, carbohydrates, and water activity. Instrumental color was influenced by the corn race effect, and the dehydration type influenced the fracturability effect. FTIR fingerprinting results revealed that hybrid samples exhibited higher intensities, particularly associated with higher lime concentrations, indicating a greater exposure of glycosidic or protein structures. Race and dehydration type effects impacted the intensity of sensory attributes, emotions, and memories. PREFMAP vector model results revealed that consumers preferred tostadas from the Solar-Chiquito, Hybrid-Pepitilla, Hybrid-Cónico, and Hybrid-Chiquito races for their higher protein content, moisture, high fracturability, crunchiness, porousness, sweetness, doughy flavor, corn flavor, and burnt flavor, while images of these tostadas evoked positive emotions (tame, adventurous, free). In contrast, the Solar-Pepitilla tostada had a lower preference because it was perceived as sour and lime-flavored, and its tostada images evoked more negative emotions and memories (worried, accident, hurt, pain, wild) and fewer positive cognitive aspects (joyful, warm, rainy weather, summer, and interested). However, the tostadas of the Solar-Cónico race were the ones that were most rejected due to their high hardness and yellow to blue tones and for evoking negative emotions (nostalgic and bored). Full article

Aflatoxin B₁ (AFB₁) contamination in corn silage poses significant risks to livestock and human health. This study developed a non-destructive detection method for AFB₁ using color-sensitive arrays (CSAs). Twenty self-developed CSAs were employed to react with samples, with reflectance spectra collected using a portable spectrometer. Spectral data were optimized through seven preprocessing methods, including Standard Normal Variate (SNV), Multiplicative Scatter Correction (MSC), first-order derivative (1st D), second-order derivative (2nd D), wavelet denoising, and their combinations. Key variables were selected using five feature selection algorithms: Competitive Adaptive Reweighted Sampling (CARS), Principal Component Analysis (PCA), Random Forest (RF), Uninformative Variable Elimination (UVE), and eXtreme Gradient Boosting (XGBoost). Five machine learning models were constructed: Light Gradient Boosting Machine (LightGBM), XGBoost, Support Vector Regression (SVR), RF, and K-Nearest Neighbor (KNN). The results demonstrated significant AFB₁-responsive characteristics in three dyes: (2,3,7,8,12,13,17,18-octaethylporphynato)chloromanganese(III) (Mn(OEP)Cl), Bromocresol Green, and Cresol Red. The combined 1st D-PCA-KNN model showed optimal prediction performance, with determination coefficient ($R_p^2$ = 0.87), root mean square error (RMSEP = 0.057), and relative prediction deviation (RPD = 2.773). This method provides an efficient solution for silage AFB₁ monitoring. Full article

As a core component of machining tools and vehicles, the load-bearing and transmission performance of rolling bearings is directly related to product processing quality and driving safety, highlighting the critical importance of fault detection. To address the nonlinearity, non-stationary modulation, and low signal-to-noise ratio (SNR) observed in bearing vibration signals, we propose a fault feature extraction method based on spectral kurtosis and Hilbert envelope demodulation. First, spectral kurtosis is employed to determine the center frequency and bandwidth of the signal adaptively, and a bandpass filter is constructed to enhance the characteristic frequency components. Subsequently, the envelope spectrum is extracted through the Hilbert transform, allowing for the precise identification of fault characteristic frequencies. In the fault diagnosis stage, a multidimensional feature vector is formed by combining the kurtosis index with the amplitude ratios of inner/outer race characteristic frequencies, and fault pattern classification is accomplished using a Least-Squares Support Vector Machine (LS-SVM). To evaluate the effectiveness of the proposed method, experiments were conducted on the bearing datasets from Case Western Reserve University (CWRU) and the Machine Failure Prevention Technology (MFPT) Society. The experimental results demonstrate that the proposed method surpasses other comparative approaches, achieving identification accuracies of 95% and 100% for the CWRU and MFPT datasets, respectively. Full article

The no-insulation (NI) technique improves the stability and defect-tolerance of high-temperature superconducting (HTS) coils by enabling current redistribution, thereby reducing the risk of quenching. NI–HTS coils are widely applied in DC systems such as high-field magnets and superconducting field coils for electric machines. However, the presence of turn-to-turn contact resistance makes current distribution uneven, rendering traditional simulation methods unsuitable. To address this, a finite element method (FEM) based on the T–A formulation is proposed. This model solves coupled equations for the magnetic vector potential (A) and current vector potential (T), incorporating turn-to-turn contact resistance and anisotropic conductivity. The thin-strip approximation simplifies second-generation HTS materials as one-dimensional conductors, and a homogenization technique further reduces computational time by averaging the properties between turns, although it may limit the resolution of localized inter-turn effects. To verify the model’s accuracy, simulation results are compared against the H formulation, distributed circuit network (DCN) model, and experimental data. The proposed T–A model accurately reproduces key transient characteristics, including magnetic field evolution and radial current distribution, in both circular and racetrack NI coils. These results confirm the model’s potential as an efficient and reliable tool for transient electromagnetic analysis of NI–HTS coils. Full article

Large language models (LLMs) have attracted a lot of attention in various fields due to their superior performance, aiming to train hundreds of millions or more parameters on large amounts of text data to understand and generate natural language. As the superior performance of LLMs becomes apparent, they are increasingly being applied to knowledge graph embedding (KGE)-related tasks to improve the processing results. Traditional KGE representation learning methods map entities and relations into a low-dimensional vector space, enabling the triples in the knowledge graph to satisfy a specific scoring function in the vector space. However, based on the powerful language understanding and semantic modeling capabilities of LLMs, which have recently been invoked to varying degrees in different types of KGE-related scenarios such as multi-modal KGE and open KGE according to their task characteristics, researchers are increasingly exploring how to integrate LLMs to enhance knowledge representation, improve generalization to unseen entities or relations, and support reasoning beyond static graph structures. In this paper, we investigate a wide range of approaches for performing LLMs-related tasks in different types of KGE scenarios. To better compare the various approaches, we summarize each KGE scenario in a classification. In the article we also discuss the applications in which the methods are mainly used and suggest several forward-looking directions for the development of this new research area. Full article

The advancement of artificial intelligence (AI) in remote sensing (RS) increasingly depends on datasets that offer rich and structured supervision beyond traditional scene-level labels. Although existing benchmarks for aerial scene classification have facilitated progress in this area, their reliance on single-class annotations restricts their application to more flexible, interpretable and generalisable learning frameworks. In this study, we introduce WHU-RS19 ABZSL: an attribute-based extension of the widely adopted WHU-RS19 dataset. This new version comprises 1005 high-resolution aerial images across 19 scene categories, each annotated with a vector of 38 features. These cover objects (e.g., roads and trees), geometric patterns (e.g., lines and curves) and dominant colours (e.g., green and blue), and are defined through expert-guided annotation protocols. To demonstrate the value of the dataset, we conduct baseline experiments using deep learning models that had been adapted for multi-label classification—ResNet18, VGG16, InceptionV3, EfficientNet and ViT-B/16—designed to capture the semantic complexity characteristic of real-world aerial scenes. The results, which are measured in terms of macro F1-score, range from 0.7385 for ResNet18 to 0.7608 for EfficientNet-B0. In particular, EfficientNet-B0 and ViT-B/16 are the top performers in terms of the overall macro F1-score and consistency across attributes, while all models show a consistent decline in performance for infrequent or visually ambiguous categories. This confirms that it is feasible to accurately predict semantic attributes in complex scenes. By enriching a standard benchmark with detailed, image-level semantic supervision, WHU-RS19 ABZSL supports a variety of downstream applications, including multi-label classification, explainable AI, semantic retrieval, and attribute-based ZSL. It thus provides a reusable, compact resource for advancing the semantic understanding of remote sensing and multimodal AI. Full article

This research article examines the CO₂ laser cutting performance of Fused Filament Fabricated Acrylonitrile Styrene Acrylate (ASA) thermoplastics by analyzing the influence of plate thickness, laser power, and cutting speed on four quality characteristics: surface roughness (Ra), top kerf width (Top KW), bottom kerf width (Bottom KW), and bottom heat-affected zone (Bottom HAZ). Forty-five experiments were conducted using five thickness levels, three power levels, and three cutting speeds. To model and predict these outputs, seven machine learning approaches were employed: Autoencoder, Autoencoder–Gated Recurrent Unit, Autoencoder–Long Short-Term Memory, Random Forest, Extreme Gradient Boosting (XGBoost), Support Vector Regression, and Linear Regression. Among them, XGBoost yielded the highest accuracy across all performance metrics. Analysis of Variance results revealed that Ra is mainly affected by plate thickness, Bottom KW by cutting speed, and Bottom HAZ by power, while Top KW is influenced by all three parameters. The study proposes an effective prediction framework using multi-output modeling and hybrid deep learning, offering a data-driven foundation for process optimization. The findings are expected to support intelligent manufacturing systems for real-time quality prediction and adaptive laser post-processing of engineering-grade thermoplastics such as ASA. This integrative approach also enables a deeper understanding of nonlinear dependencies in laser–material interactions. Full article