Search Results (967)

The Internet of Things (IoT) is increasingly deployed at the edge under resource and environmental constraints, which limits the practicality of traditional intrusion detection systems (IDSs) on IoT hardware. This paper presents two IDS configurations. First, we develop a baseline IDS with fixed hyperparameters, achieving 99.20% accuracy and ~0.002 ms/sample inference latency on a desktop machine; this configuration is suitable for high-performance platforms but is not intended for constrained IoT deployment. Second, we propose a lightweight, edge-oriented IDS that applies ANOVA-based filter feature selection and uses a genetic algorithm (GA) for the bounded hyperparameter tuning of the classifier under stratified cross-validation, enabling efficient execution on Raspberry Pi-class devices. The lightweight IDS achieves 98.95% accuracy with ~4.3 ms/sample end-to-end inference latency on Raspberry Pi while detecting both low-volume and high-volume (DoS/DDoS) attacks. Experiments are conducted in a Raspberry Pi-based real lab using an up-to-date mixed-modal dataset combining system/network telemetry and heterogeneous physical sensors. Overall, the proposed framework demonstrates a practical, hardware-aware, and reproducible way to balance detection performance and edge-level latency using established techniques for real-world IoT IDS deployment. Full article

Avobenzone (AVO) and ethylhexyl salicylate (EHS) are widely used organic UV filters with distinct photochemical properties and reported biological effects. Experimental and predictive evidence suggests that some lipophilic UV filters may reach systemic circulation and potentially cross the blood–brain barrier (BBB), raising concerns about possible central nervous system effects, although direct evidence for AVO and EHS remains limited. This study evaluated the effects of subcytotoxic concentrations (0.01–1 µM) of AVO and EHS on differentiated SH-SY5Y human neuroblastoma cells, focusing on early stress-related molecular responses. Cell viability and reactive oxygen species production were not significantly affected at any tested concentration. Integrated analyses of microRNA, gene, and protein expression revealed modest and variable modulation of miR-200a-3p and miR-29b-3p. Western blot analysis showed increased phosphorylation of AKT and ERK, no significant changes in mTOR activation, and an increased Bax/Bcl-2 ratio. Overall, these findings indicate that AVO and EHS trigger an early stress-adaptive response involving PI3K/AKT and MAPK/ERK signaling and modulation of apoptosis-related pathways. Such responses reflect a dynamic balance between cellular adaptation and pro-apoptotic signaling, which may become relevant under prolonged or higher-intensity exposure conditions. Full article

Collateral status is an important therapeutic indicator for acute ischemic stroke (AIS), yet visual collateral grading remains subjective and suffers from inter-observer variability. To address this limitation, this study automatically extracted binarized vascular morphological features from CTA images and developed a convolutional neural network (CNN) for automated collateral classification. Performance trends were systematically analyzed across diverse hyperparameter combinations to meet different clinical decision needs. A total of 157 AIS patients (median age 65 [57–74] years; 61.8% were male) were retrospectively enrolled and stratified by Menon score into good (3–5, n = 117) and poor (0–2, n = 40) collateral groups. A total of 192 architectures were established, and three representative model tendencies emerged: a sensitivity-oriented model (AUC = 0.773; sensitivity = 87.18%; specificity = 65.00%), a balanced model (AUC = 0.768; sensitivity = 72.65%; specificity = 77.50%), and a specificity-oriented model (AUC = 0.753; sensitivity = 63.25%; specificity = 85.00%). These results demonstrate that kernel size, the number of filters in the first layer, and the number of convolutional layers are key determinants of performance directionality, allowing tailored model selection depending on clinical requirements. This work highlights the feasibility of CTA-based automated collateral classification and provides a systematic framework for developing models optimized for sensitivity, specificity, or balanced decision-making. The findings may serve as a reference for clinical model deployment and have potential for integration into multi-objective AI systems for endovascular thrombectomy patient triage. Full article

Accurate and interpretable detection of pulmonary emphysema on low-dose computed tomography (LDCT) remains a critical challenge for large-scale screening and population health studies. This work proposes a quality-controlled and interpretable deep learning pipeline for emphysema assessment using ResNet-152 embeddings. The pipeline integrates automated lung segmentation, quality-control filtering, and extraction of 2048-dimensional embeddings from mid-lung patches, followed by analysis using logistic regression, LASSO, and recursive feature elimination (RFE). The embeddings are further fused with quantitative CT (QCT) markers, including %LAA, Perc15, and total lung volume (TLV), to enhance robustness and interpretability. Bootstrapped validation demonstrates strong diagnostic performance (ROC-AUC = 0.996, PR-AUC = 0.962, balanced accuracy = 0.931) with low computational cost. The proposed approach shows that ResNet embeddings pretrained on CT data can be effectively reused without retraining for emphysema characterization, providing a reproducible and explainable framework suitable as a research and screening-support framework for population-level LDCT analysis. Full article

Within the domain of intelligent picking robotics, fruit recognition and positioning are essential. Challenging conditions such as varying light, occlusion, and limited edge-computing power compromise fruit maturity detection. To tackle these issues, this paper proposes a lightweight algorithm YOLO-ELS based on YOLOv8n. Specifically, we reconstruct the backbone by replacing the bottlenecks in the C2f structure with Edge-Information-Enhanced Modules (EIEM) to prioritize morphological cues and filter background redundancy. Furthermore, a Large Separable Kernel Attention (LSKA) mechanism is integrated into the SPPF layer to expand the effective receptive field for multi-scale targets. To mitigate occlusion-induced errors, a Spatially Enhanced Attention Module (SEAM) is incorporated into the decoupled detection head to enhance feature responses in obscured regions. Finally, the Inner-GIoU loss is adopted to refine bounding box regression and accelerate convergence. Experimental results demonstrate that compared to the YOLOv8n baseline, the proposed YOLO-ELS achieves a 14.8% reduction in GFLOPs and a 2.3% decrease in parameters, while attaining a precision, recall, and mAP@50% of 92.7%, 83.9%, and 92.0%, respectively. When compared with mainstream models such as DETR, Faster-RCNN, SSD, TOOD, YOLOv5s, and YOLO11n, the mAP@50% is improved by 7.0%, 4.7%, 11.4%, 8.6%, 3.1%, and 3.2%. Deployment tests on the NVIDIA Jetson Orin Nano Super edge platform yield an inference latency of 25.2 ms and a detection speed of 28.2 FPS, successfully meeting the real-time operational requirements of automated harvesting systems. These findings confirm that YOLO-ELS effectively balances high detection accuracy with lightweight architecture, providing a robust technical foundation for intelligent fruit picking in resource-constrained greenhouse environments. Full article

With the increasing integration of renewable energy sources into the grid, power quality at the point of common coupling (PCC)—particularly harmonic distortion introduced by power electronic converters—has become a critical concern. This paper presents a rigorous design and evaluation of a three-phase, fifteen-level cascaded H-bridge multilevel inverter (CHB MLI) with an LCL filter, selected for its superior harmonic attenuation, compact size, and cost-effectiveness compared to conventional passive filters. The proposed system employs Phase-Shifted Pulse Width Modulation (PS PWM) for balanced operation and low output distortion. A systematic, reproducible methodology is used to design the LCL filter, which is then tested across a wide range of switching frequencies (1–5 kHz) and grid impedance ratios (X/R = 2–9) in MATLAB/Simulink R2025a. Comprehensive simulations confirm that the filter effectively reduces both voltage and current total harmonic distortion (THD) to levels well below the 5% limit specified by IEEE 519, with optimal performance (0.53% current THD, 0.69% voltage THD) achieved at 3 kHz and X/R ≈ 5.6. The filter demonstrates robust performance regardless of grid conditions, making it a practical and scalable solution for modern renewable energy integration. These results, further supported by parametric validation and clear design guidelines, provide actionable insights for academic research and industrial deployment. Full article

Adaptive filtering remains a cornerstone of modern signal processing but faces fundamental challenges when confronted with rapidly changing or nonlinear environments. This work investigates the integration of deep learning into adaptive-filter architectures to enhance tracking capability and robustness in non-stationary conditions. After reviewing and analyzing classical algorithms—LMS, NLMS, RLS, and a variable step-size LMS (VSS-LMS)—their theoretical stability and mean-square error behavior are formalized under a slow-variation system model. Comprehensive simulations using drifting autoregressive (AR(2)) processes, piecewise-stationary FIR systems, and time-varying sinusoidal signals confirm the classical trade-off between performance and complexity: RLS achieves the lowest steady-state error, at a quadratic cost, whereas LMS remains computationally efficient with slower adaptation. A stabilized VSS-LMS algorithm is proposed to balance these extremes; the results show that it maintains numerical stability under abrupt parameter jumps while attaining steady-state MSEs that are comparable to RLS (approximately 3 × 10⁻²) and superior robustness to noise. These findings are validated by theoretical tracking-error bounds that are derived for bounded parameter drift. Building on this foundation, a deep-learning-driven adaptive filter is introduced, where the update rule is parameterized by a neural function, $U_{\theta }$, that generalizes the classical gradient descent. This approach offers a pathway toward adaptive filters that are capable of self-tuning and context-aware learning, aligning with emerging trends in AI-augmented system architectures and next-generation computing. Future work will focus on online learning and FPGA/ASIC implementations for real-time deployment. Full article

Widely used in treating oil mist aerosols generated from metalworking processes, conventional gas–liquid coalescing filters face drawbacks such as increased energy consumption, performance limitations, and shortened service life due to high steady-state pressure drop. To address these issues, this study proposes an innovative design for a filter based on wettability-regulated gradient pore structure. Using glass fiber filter media with different pore size parameters as the substrate and incorporating an intermediate mesh layer, a three-layer filtration structure of “large-pore filtration layer—mesh layer—small-pore filtration layer” was constructed. The surface wettability of each layer was regulated by a self-developed surface modifier, producing gradient pore structure filters with different wettability configurations. The variations in key performance parameters, including steady-state pressure drop, filtration efficiency, saturation, and service life, were systematically evaluated for these configurations. Experimental results demonstrated that the configuration with an “oleophobic large-pore filtration layer—mesh layer—oleophilic small-pore filtration layer” yielded the best overall performance. Analysis based on the “jump-channel” model indicated that the gradient pore structure achieves progressive droplet filtration and optimizes droplet coalescence and capture through wettability differences. Consequently, while maintaining exceptional filtration efficiency (>99%), this configuration significantly reduces the steady-state pressure drop by over 34% and effectively extends the service life by more than 66%. This wettability-regulated gradient pore structure provides a novel technical pathway for addressing the challenges of balancing pressure drop and filtration efficiency, as well as extending the service life, in gas–liquid coalescing filters. Full article

Endometrial lesions are common in gynecology, exhibiting considerable clinical heterogeneity across different subtypes. Although ultrasound imaging is the preferred diagnostic modality due to its noninvasive, accessible, and cost-effective nature, its diagnostic performance remains highly operator-dependent, leading to subjectivity and inconsistent results. To address these limitations, this study proposes a hierarchical feature selection and optimization framework for endometrial lesions, aiming to enhance the objectivity and robustness of ultrasound-based diagnosis. Firstly, Kernel Principal Component Analysis (KPCA) is employed for nonlinear dimensionality reduction, retaining the top 1000 principal components. Secondly, an ensemble of three filter-based methods—information gain, chi-square test, and symmetrical uncertainty—is integrated to rank and fuse features, followed by thresholding with Maximum Scatter Difference Linear Discriminant Analysis (MSDLDA) for preliminary feature selection. Finally, the Whale Migration Algorithm (WMA) is applied to population-based feature optimization and classifier training under the constraints of a Support Vector Machine (SVM) and a macro-averaged F1 score. Experimental results demonstrate that the proposed closed-loop pipeline of “kernel reduction—filter fusion—threshold pruning—intelligent optimization—robust classification” effectively balances nonlinear structure preservation, feature redundancy control, and model generalization, providing an interpretable, reproducible, and efficient solution for intelligent diagnosis in small- to medium-scale medical imaging datasets. Full article

Many denoising methods model noise as Gaussian noise. However, the realistic noise captured by camera devices does not satisfy Gaussian distribution. Hence, those methods do not perform well when being applied to real-world image denoising tasks. In this work, we indicate that the spatial correlation in noise and the variation of noise intensity are the main factors that impact the performance of Gaussian noise-based methods, and accordingly propose an extension of the method based on the weighted, mixed non-convex ${\mathsf{\ell }}_p$ norm. The proposed method first strengthens the intensity of the noise pattern in the original denoising result through the Guided Filter, then removes the over-amplified frequency in the local area by the proposed regularization term. We prove that the optimal solution can be achieved through the sub-gradient-based iterative optimization scheme, and further reduce the computational cost by optimizing the initial values. Numerical experiments show that the proposed extending method can balance well texture preservation and noise removal, and the PSNR of the extending method’s results are greatly improved, even outperforming the recently proposed realistic noise removal methods which also include deep learning based methods. Full article

This study proposes an explainability-driven feature selection framework for phishing website detection using a large-scale, heterogeneous dataset collected from four independent sources. The combined dataset contains approximately 500,000 samples, including 300,000 phishing pages and 200,000 legitimate pages, providing a comprehensive representation of real-world web traffic. To enhance model interpretability and reduce feature redundancy, four explainable artificial intelligence (XAI) techniques—SHAP, LIME, partial dependence plots (PDPs), and permutation importance (PDI)—were applied to rank and analyze feature contributions. The union of all selected features was subsequently refined through a thresholding mechanism, forming the proposed Hybrid Explainability Random Forest Algorithm (HXRF). A Random Forest (RF) classifier was trained using the optimized feature subset and evaluated on an independently sampled set of 2000 webpages. Results demonstrate that HXRF significantly improves classification performance, achieving an accuracy of 98.2%, with balanced precision, recall, and F1 scores. The confusion matrix confirms strong generalization across both phishing and legitimate classes, with minimal false predictions. This work demonstrates that combining multi-method XAI with selective feature filtering produces a compact, interpretable, and highly discriminative feature set capable of robust phishing detection at scale. Full article

Understanding the spatial structure of large mammals is critical for conservation planning, especially under increasing habitat fragmentation. This study applies an integrated spatial analysis combining the DBSCAN density-based clustering algorithm and the α-hull method to delineate non-convex geographic ranges of the maned wolf (Chrysocyon brachyurus) across South America. Using 454 occurrence records filtered for ecological reliability, we identified 11 geographically isolated α-populations distributed across five countries and multiple biomes, including the Cerrado, Chaco, and Atlantic Forest. The sensitivity analysis of the α parameter demonstrated that values below 2 failed to generate viable polygons, while α = 2 provided the best balance between geometric detail and ecological plausibility. Our results reveal a highly fragmented distribution, with α-populations varying in area from 43,077 km² to 566,154.7 km² and separated by distances up to 994.755 km. Smaller and peripheral α-populations are likely more vulnerable to stochastic processes, genetic drift, and inbreeding, while larger clusters remain functionally isolated due to anthropogenic barriers. We propose the concept of ‘α-population’ as an operational unit to describe geographically and functionally isolated groups identified through combined spatial clustering and non-convex hull analysis. This approach offers a reproducible and biologically meaningful framework for refining range estimates, identifying conservation units, and guiding targeted management actions. Overall, integrating α-hulls with density-based clustering improves our understanding of the species’ fragmented spatial structure and supports evidence-based conservation strategies aimed at maintaining habitat connectivity and long-term viability of C. brachyurus populations. Full article

In this work, a substrate-integrated waveguide (SIW) filtering power divider with a modified complementary split-ring resonator (CSRR) is reported. Firstly, by integrating the meander-shaped slots with the conventional CSRR, the proposed inner-meander-slot CSRR (IMSCSRR) can enlarge the total length of the defected slot and increase the width of the split, thus enhancing the equivalent capacitance and inductance. In this way, the fundamental resonant frequency of the IMSCSRR can be effectively decreased without enlarging the circuit size, which can generally help to reduce the physical size by over 35%. Subsequently, to further reduce the circuit size, two IMSCSRRs are separately loaded on the top and bottom metal covers to constitute a broadside-coupled IMSCSRR, which is combined with the SIW. To verify the efficacy of the proposed SIW-IMSCSRR unit cell, a two-way filtering power divider is implemented. It combines the band-selection function of a filter and the power-distribution property of a power divider, thereby enhancing system integration and realizing size compactness. Experimental results show that the proposed filtering power divider achieves a center frequency of 3.53 GHz, a bandwidth of about 320 MHz, an in-band insertion loss of (3 + 1.3) dB, an in-band isolation of over 21 dB, and a size reduction of about 30% compared with the design without broadside-coupling, as well as good magnitude and phase variations. All the results indicate that the proposed filtering power divider achieves a good balance between low loss, high isolation, and compact size, which is suitable for system integration applications in microwave scenarios. Full article

This study presents an analysis of the Infrastructure Leakage Index (ILI) variability for two District Metered Areas (DMAs) in the Silesian Region (Poland), based on 2024 data. The objective of the study was to evaluate whether high-frequency AMI data can be used to reliably identify and remove distorted measurement periods, thereby improving the credibility of the annual ILI value for each individual DMA. ILI_T values were calculated for daily, weekly, and monthly intervals using synchronized hourly data from an Advanced Metering Infrastructure (AMI) system and water network monitoring platforms. A key methodological advantage was the use of fully synchronous inflow–outflow–consumption data, enabling diagnostic reconstruction of hourly water balances and validation of the representativeness of data segments used for ILI_T estimation. The study applied statistical measures of variability (standard deviation, variance, coefficient of variation) and graphical methods (histograms, boxplots) to evaluate ILI_T behavior across time resolutions. Rather than comparing leakage performance between DMAs—which is performed exclusively using normalized indicators such as ILI—the analysis examined how hourly diagnostic information explains short-term distortions in the ILI and how filtering such periods affects the stability of the annual value for each DMAs. The results confirm that ILI_T interpretation is highly dependent on temporal resolution. Daily data is more responsive to anomalies and operational events, while monthly data provides more stable values suitable for benchmarking. The findings demonstrate that daily and hourly data should be used diagnostically to detect non-representative periods, whereas monthly aggregation provides the most robust basis for reporting and inter-DMA comparison. Overall, the study proposes a practical procedure for ILI validation using AMI data and demonstrates its application on two real DMAs. Full article

Electromagnetic flowmeters face significant challenges in measuring complex fluids, characterized by weak flow signals and severe noise interference. Conventional solutions, such as dual-frequency rectangular wave excitation, suffer from multiple drawbacks including rich harmonic components, high electromagnetic noise during switching transitions, a propensity for resonance which shortens stabilization time, reduced sampling windows, and complex circuit implementation. Similarly, traditional single-frequency excitation struggles to balance zero stability with the suppression of slurry noise. To address these limitations, this paper proposes a novel converter design based on dual-frequency sinusoidal wave excitation. A pure hardware circuit is used to generate the composite excitation signal, which superimposes low-frequency and high-frequency components. This approach eliminates the need for a master control chip in signal generation, thereby reducing both circuit complexity and computational resource allocation. The signal processing chain employs a technique of “high-order Butterworth separation filtering combined with synchronous demodulation,” effectively suppressing power frequency, orthogonal, and in-phase interference, achieving an improvement in interference rejection by approximately three orders of magnitude (1000×). Experimental results show that the proposed converter featured simplified circuitry, achieved a measurement accuracy of class 0.5, and validated the overall feasibility of the scheme. Full article