Search Results (16,632)

Non-Intrusive Load Monitoring is an effective method for disaggregating the power consumption of individual appliances from the aggregate load data of a building. The advent of smart meters, Internet of Things devices, and artificial intelligence technologies has significantly advanced the capabilities of non-intrusive load monitoring. However, challenges such as varying sampling frequencies and measurement sensitivities remain. This paper introduces an innovative model incorporating an Adaptive Multi-Scale Attention Integration Module (AMSAIM) to address these issues. The model leverages deep learning and attention mechanisms to improve the accuracy and real-time performance of non-intrusive load monitoring. Validated on the standard UK-DALE dataset, the model consistently demonstrated superior performance. In seen scenarios, our model achieved average F1-scores approximating 0.94 and notably reduced Mean Absolute Error (MAE) values. For washing machines, it achieved an F1-score of 0.99 and MAE of 41.64, outperforming the next best method’s F1-score by 1 percentage point. In challenging unseen scenarios, the model showcased strong generalization, achieving an F1-score of 0.91 for washing machines and reducing MAE to 7.66. Furthermore, an ablation study rigorously confirmed the necessity of the AMSAIM module, showing that the synergistic integration of the efficient multi-scale attention (EMA) and the selective kernel (SK) adaptive receptive field unit is crucial for enhancing model robustness and generalization. Our results highlight the model’s potential for enhancing energy efficiency and providing actionable insights for energy management across various conditions. Full article

In nanometer integrated-circuit (IC) manufacturing, advanced technology scaling has intensified the effects of process variations on circuit reliability and performance. Random fluctuations in parameters such as threshold voltage, channel length, and oxide thickness further degrade design margins and increase the likelihood of functional failures. These variations often lead to rare circuit failure events, underscoring the importance of accurate yield estimation and robust design methodologies. Conventional Monte Carlo yield estimation is computationally infeasible as millions of simulations are required to capture failure events with extremely low probability. This paper presents a novel reliability-based circuit design optimization framework that leverages deep transfer learning to improve the efficiency of repeated yield analysis in optimization iterations. Based on pre-trained neural network models from prior design knowledge, we utilize model fine-tuning to accelerate importance sampling (IS) for yield estimation. To improve estimation accuracy, adversarial perturbations are introduced to calibrate uncertainty near the model decision boundary. Moreover, we propose a cascaded batch Bayesian optimization (CBBO) framework that incorporates a smart initialization strategy and a localized penalty mechanism, guiding the search process toward high-yield regions while satisfying nominal performance constraints. Experimental validation on SRAM circuits and amplifiers reveals that CBBO achieves a computational speedup of 2.02×–4.63× over state-of-the-art (SOTA) methods, without compromising accuracy and robustness. Full article

Ensuring sustained driver attention is critical for intelligent transportation safety systems; however, the performance of data-driven driver distraction detection models is often limited by the high cost of large-scale manual annotation. To address this challenge, this paper proposes an adaptive semi-supervised driver distraction detection framework based on teacher–student learning and deformable pyramid feature fusion. The framework leverages a limited amount of labeled data together with abundant unlabeled samples to achieve robust and scalable distraction detection. An adaptive pseudo-label optimization strategy is introduced, incorporating category-aware pseudo-label thresholding, delayed pseudo-label scheduling, and a confidence-weighted pseudo-label loss to dynamically balance pseudo-label quality and training stability. To enhance fine-grained perception of subtle driver behaviors, a Deformable Pyramid Sparse Transformer (DPST) module is integrated into a lightweight YOLOv11 detector, enabling precise multi-scale feature alignment and efficient cross-scale semantic fusion. Furthermore, a teacher-guided feature consistency distillation mechanism is employed to promote semantic alignment between teacher and student models at the feature level, mitigating the adverse effects of noisy pseudo-labels. Extensive experiments conducted on the Roboflow Distracted Driving Dataset demonstrate that the proposed method outperforms representative fully supervised baselines in terms of mAP@0.5 and mAP@0.5:0.95 while maintaining a balanced trade-off between precision and recall. These results indicate that the proposed framework provides an effective and practical solution for real-world driver monitoring systems under limited annotation conditions. Full article

The V/G criterion is a critical indicator for monitoring dynamic changes during Czochralski silicon single crystal (Cz-SSC) growth. However, the inability to measure it in real time forces reliance on offline feedback for process regulation, leading to imprecise control and compromised crystal quality. To overcome this limitation, this paper proposes a novel soft sensor modeling framework that integrates both mechanism-based knowledge and data-driven learning for the real-time prediction of the crystal quality parameter, specifically the V/G value (the ratio of growth rate to axial temperature gradient). The proposed approach constructs a hybrid prediction model by combining a data-driven sub-model with a physics-informed mechanism sub-model. The data-driven component is developed using an attention-based dynamic stacked enhanced autoencoder (AD-SEAE) network, where the SEAE structure introduces layer-wise reconstruction operations to mitigate information loss during hierarchical feature extraction. Furthermore, an attention mechanism is incorporated to dynamically weigh historical and current samples, thereby enhancing the temporal representation of process dynamics. In addition, a robust ensemble approach is achieved by fusing the outputs of two subsidiary models using an adaptive weighting strategy based on prediction accuracy, thereby enabling more reliable V/G predictions under varying operational conditions. Experimental validation using actual industrial Cz-SSC production data demonstrates that the proposed method achieves high-prediction accuracy and effectively supports real-time process optimization and quality monitoring. Full article

Karst aquifers provide critical water resources in the Mediterranean region, yet climate change threatens their sustainability. This study integrates stable isotope analysis (δ²H, δ¹⁸O), hydrochemistry, and hydrological time series to characterize precipitation–groundwater dynamics in the Mt. Učka karst system (Croatia). Precipitation samples collected across an altitudinal gradient of approximately 1400 m and groundwater from three major groundwater sources were analyzed over a 2.5-year period. Precipitation exhibits pronounced isotopic variability with d-excess values indicating mixed Atlantic–Mediterranean moisture sources. Groundwater is primarily recharged by precipitation from the cold part of the hydrological year. It exhibits substantial attenuation of isotopic signals, which indicates extensive mixing processes but prevents quantitative estimation of mean residence time. Groundwater is predominantly recharged from elevations above 900 m a.s.l., with one spring showing evidence of higher-elevation recharge. Analysis confirms the system’s dual porosity: a rapid, conduit-dominated response indicates high vulnerability to surface contamination, while a sustained, matrix-dominated response provides greater buffering capacity. These findings highlight the vulnerability of karst systems to projected reductions in autumn precipitation, the critical recharge season, and demonstrate the necessity of multi-tracer approaches for comprehensive aquifer characterization. Full article

Owing to their high strength characteristics, chemical stability, and piezoelectric activity, vinylidene fluoride (VDF) copolymers have become promising materials for creating implants to replace bone tissue defects. However, a significant drawback of these materials is the biological inertness of their surface, which leads to unsatisfactory integration with the patient’s bone tissue. In this study, we propose a single-step approach for immobilizing hydroxyapatite (HAp) on the surface of porous implants made of vinylidene fluoride and tetrafluoroethylene copolymer (P(VDF-TeFE)). This method consists of treating the surface of the product with a mixture of solvents while simultaneously capturing HAp microparticles. Using scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS), it was shown that the proposed method preserves the morphology of model implants (pore diameter and printed line thickness) and allows HAp to cover up to 63 ± 14% of their surface, reaching concentrations of calcium and phosphorus up to 6.0 ± 1.3 and 3.6 ± 0.7 at. %, respectively, imparting superhydrophilic properties to them. Optical profilometry revealed that the surface roughness of samples increased by more than seven times as a result of HAp immobilization. X-ray diffraction analysis (XRD) confirmed that the piezoelectric phase of P(VDF-TeFE) is preserved after treatment, as are the compressive strength characteristics of the samples. Hydroxyapatite immobilization significantly improved the adhesion and osteogenic differentiation of multipotent stem cells cultured with P(VDF-TeFE)-based samples. Thus, the proposed method can significantly enhance the biological activity of implants based on the piezoelectric VDF copolymer. Full article

The rapid advancement of sequencing technologies has led to the generation of complex multi-omics data, which are often high-dimensional, noisy, and imbalanced, posing significant challenges for traditional machine learning methods. The novelty of this work resides in the architecture-level integration of autoencoders with Generative Adversarial Network (GAN) and Conditional Tabular Generative Adversarial Network (CTGAN) models, where the autoencoder is employed for latent feature extraction and noise reduction, while GAN-based models are used for realistic sample generation and class imbalance mitigation in multi-omics cancer datasets. This study proposes a novel framework that combines an autoencoder for dimensionality reduction and a CTGAN for generating synthetic samples to balance underrepresented classes. The process starts with selecting the most discriminative features, then extracting latent representations for each omic type, merging them, and generating new minority samples. Finally, all samples are used to train a neural network to predict specific cancer outcomes, defined here as clinically relevant biomarkers or patient characteristics. In this work, the considered outcome in the bladder cancer is Tumor Mutational Burden (TMB), while the breast cancer outcome is menopausal status, a key factor in treatment planning. Experimental results show that the proposed model achieves high precision, with an average precision of 0.9929 for TMB prediction in bladder cancer and 0.9748 for menopausal status in breast cancer, and reaches perfect precision (1.000) for the positive class in both cases. In addition, the proposed AE–CTGAN framework consistently outperformed an autoencoder combined with a standard GAN across all evaluation metrics, achieving average accuracies of 0.9929 and 0.9748, recall values of 0.9846 and 0.9777, and F1-scores of 0.9922 for bladder and breast cancer datasets, respectively. A comparative fidelity analysis in the latent space further demonstrated the superiority of CTGAN, reducing the average Euclidean distance between real and synthetic samples by approximately 72% for bladder cancer and by up to 84% for breast cancer compared to a standard GAN. These findings confirm that CTGAN generates high-fidelity synthetic samples that preserve the structural characteristics of real multi-omics data, leading to more reliable class balancing and improved predictive performance. Overall, the proposed framework provides an effective and robust solution for handling class imbalance in multi-omics cancer data and enhances the accuracy of clinically relevant outcome prediction. Full article

This study investigates whether fecal microbiota transplantation (FMT) can alleviate gut microbiota dysbiosis induced by a high-fat diet (HFD) through modulation of fatty acid metabolism, competition for nutrients, production of short-chain fatty acids (SCFAs), and restoration of mucus layer integrity. To elucidate the mechanisms by which FMT regulates colonic microbial function and host metabolic responses, 80 male Bal b/c mice were randomly assigned to four experimental groups (n = 20 per group): Normal Diet Group (NDG), High-Fat Diet Group (HDG), Restrictive Diet Group (RDG), and HDG recipients of NDG-derived fecal microbiota (FMT group). The intervention lasted for 12 weeks, during which body weight was monitored biweekly. At the end of the experiment, tissue and fecal samples were collected to assess digestive enzyme activities, intestinal histomorphology, gene expression related to gut barrier function, and gut microbiota composition via 16S rRNA gene sequencing. Results showed that mice in the HDG exhibited significantly higher final body weight and greater weight gain compared to those in the NDG and RDG (p < 0.05). Notably, FMT treatment markedly attenuated HFD-induced weight gain (p < 0.05), reducing it to levels comparable with the NDG (p > 0.05). While HFD significantly elevated the activities of α-amylase and trypsin (p < 0.05), FMT supplementation effectively suppressed these enzymatic activities (p < 0.05). Moreover, FMT ameliorated HFD-induced intestinal architectural damage, as evidenced by significant increases in villus height and the villus height-to-crypt depth ratio (V/C) (p < 0.05). At the molecular level, FMT significantly downregulated the expression of pro-inflammatory cytokines (IL-1β, IL-1α, TNF-α) and upregulated key tight junction proteins (Occludin, Claudin-1, ZO-1) and mucin-2 (MUC2) relative to the HDG (p < 0.05). 16S rRNA analysis demonstrated that FMT substantially increased the abundance of beneficial genera such as Lactobacillus and Bifidobacterium while reducing opportunistic pathogens including Romboutsia (p < 0.05). Furthermore, alpha diversity indices (Chao1 and ACE) were significantly higher in the FMT group than in all other groups (p < 0.05), indicating enhanced microbial richness and community stability. Functional prediction using PICRUSt2 revealed that FMT-enriched metabolic pathways (particularly those associated with SCFA production) and enhanced gut barrier-related functions. Collectively, this study deepens our understanding of host–microbe interactions under HFD-induced metabolic stress and provides mechanistic insights into how FMT restores gut homeostasis, highlighting its potential as a therapeutic strategy for diet-induced dysbiosis and associated metabolic disorders. Full article

The current intrusion detection methods suffer from deficiencies in terms of cross-domain adaptability, privacy preservation, and limited effectiveness in detecting minority-class attacks. To address these issues, a novel intrusion detection model framework, TrMulS, is proposed that integrates federated learning, generative adversarial networks with multispace feature enhancement ability, and transformers with multi-source transfer ability. First, at each institution (source domain), local spatial features are extracted through a CNN, multiple subsets are constructed (to solve the feature singularity problem), and the multihead self-attention mechanism of the transformer is utilized to capture the correlation of features. Second, the synthetic samples of the target domain are generated on the basis of the improved Exchange-GAN, and the cross-domain transfer module is designed by combining the Maximum Mean Discrepancy (MMD) to minimize the feature distribution difference between the source domain and the target domain. Finally, the federated transfer learning strategy is adopted. The model parameters of each local institution are encrypted and uploaded to the target server and then aggregated to generate the global model. These steps iterate until convergence, yielding the globally optimal model. Experiments on the ISCX2012, KDD99 and NSL-KDD intrusion detection standard datasets show that the detection accuracy of this method is significantly improved in cross-domain scenarios. This paper presents a novel paradigm for cross-domain security intelligence analysis that considers efficiency, privacy and balance. Full article

Conventional descriptive statistical approaches in per- and polyfluoroalkyl substance (PFAS) environmental forensics often fail under small-sample, ecosystem-level complexity, challenging the optimization of sampling, monitoring, and remediation strategies. This study presents an advance from passive description to adaptive decision-support for complex PFAS contamination. By integrating Bayesian optimization (BO) via Gaussian Processes (GP) with a Generalized Linear Mixed Model (GLMM), we developed a signal-extraction framework for both understanding and action from limited data (n = 18). The BO/GP model achieved strong predictive performance (GP leave-one-out R² = 0.807), while the GLMM confirmed significant overdispersion (1.62), indicating a patchy contamination distribution. The integrated analysis suggested a dominant spatiotemporal interaction: a transient, high-intensity perfluorooctane sulfonate (PFOS) plume that peaked at a precise location during early November (the autumn recharge period). Concurrently, the GLMM identified significant intra-sample variance (p = 0.0186), suggesting likely particulate-bound (colloid/sediment) transport, and detected n-ethyl perfluorooctane sulfonamidoacetic acid (NEtFOSAA) as a critical precursor (p < 0.0001), thus providing evidence consistent with the source as historic 3M aqueous film-forming foam. This coupled approach creates a dynamic, iterative decision-support system where signal-based diagnosis informs adaptive optimization, enabling mission-specific actions from targeted remediation to monitoring design. Full article

Accurate displacement field measurement by holographic interferometry requires robust analysis of high-density fringe patterns, which is hindered by speckle noise inherent in any interferogram, no matter how perfect. Conventional skeletonization methods, such as edge detection algorithms and active contour models, often fail under these conditions, producing fragmented and unreliable fringe contours. This paper presents a novel skeletonization procedure that simultaneously addresses three fundamental challenges: (1) topology preservation—by representing the fringe family within a physics-informed, finite-dimensional parametric subspace (e.g., Fourier-based contours), ensuring global smoothness, connectivity, and correct nesting of each fringe; (2) extreme noise robustness—through a robust strip integration functional that replaces noisy point sampling with Gaussian-weighted intensity averaging across a narrow strip, effectively suppressing speckle while yielding a smooth objective function suitable for gradient-based optimization; and (3) sub-pixel accuracy without phase extraction—leveraging continuous bicubic interpolation within a recursive quasi-optimization framework that exploits fringe similarity for precise and stable contour localization. The method’s performance is quantitatively validated on synthetic interferograms with controlled noise, demonstrating significantly lower error compared to baseline techniques. Practical utility is confirmed by successful processing of a real interferogram of a bent plate containing over 100 fringes, enabling precise displacement field reconstruction that closely matches independent theoretical modeling. The proposed procedure provides a reliable tool for processing challenging interferograms where traditional methods fail to deliver satisfactory results. Full article

This paper proposes an event-triggered extension of duty-ratio-based model predictive direct speed control (DR-MPDSC) for permanent magnet synchronous motor (PMSM) drives in electric vehicle (EV) applications. The main contribution is the development of an event-triggered execution framework specifically tailored to DR-MPDSC, in which control updates are performed only when the speed tracking error violates a prescribed condition, rather than at every sampling instant. Unlike conventional MPDSC and time-triggered DR-MPDSC schemes, the proposed strategy achieves a significant reduction in control execution frequency while preserving fast dynamic response and closed-loop stability. An optimized duty-ratio formulation is employed to regulate the effective application duration of the selected voltage vector within each sampling interval, resulting in reduced electromagnetic torque ripple and improved stator current quality. An extended Kalman filter (EKF) is integrated to estimate rotor speed and load torque, enabling disturbance-aware predictive speed control without mechanical torque sensing. Furthermore, a unified field-weakening strategy is incorporated to ensure wide-speed-range operation under constant power constraints, which is essential for EV traction systems. Simulation and experimental results demonstrate that the proposed event-triggered DR-MPDSC achieves steady-state speed errors below 0.5%, limits electromagnetic torque ripple to approximately 2.5%, and reduces stator current total harmonic distortion (THD) to 3.84%, compared with 5.8% obtained using conventional MPDSC. Moreover, the event-triggered mechanism reduces control update executions by up to 87.73% without degrading transient performance or field-weakening capability. These results confirm the effectiveness and practical viability of the proposed control strategy for high-performance PMSM drives in EV applications. Full article

Land-use conversion from forest-to-pasture in the Amazon can affect soil physical quality and hydraulic functioning. The study evaluates the effects of land use (forest and pasture) and soil texture (fine and coarse) on soil structure and hydraulic properties, using the soil water retention curve as an integrative indicator. The study was conducted with soil samples from the Tapajós National Forest region, Pará State, Brazil, with eight sites (four forest and four pasture), balanced by texture. Undisturbed samples were collected from five profile layers (0–10, 10–20, 20–30, and 30–40 cm) for each site, totaling 160 samples. Samples were saturated and measured at soil water matric potentials from −0.1 to −15,000 hPa to obtain the soil water retention curve, which was fitted using the van Genuchten–Mualem model. Pore size distribution was derived from the relationship between soil water matric potential and equivalent pore diameter. Results are reported for the 0–40 cm soil profile (integrating the four sampled layers). Forest-to-pasture conversion altered soil pore structure and water retention in a texture-dependent manner. For fine-textured soils, bulk density increased from 1.03 to 1.31 Mg m⁻³ (+27%) from forest to pasture. In coarse-textured soils, the drainable pore volume up to −15,000 hPa, equivalent diameter > 0.2 µm) decreased from 0.296 to 0.147 m³ m⁻³ (−50%) from forest to pasture. Plant-available water across the 0–40 cm profile ranged from 0.107 m³ m⁻³ (pasture, fine-textured) to 0.137 m³ m⁻³ (forest, coarse-textured). Coarse-textured soils showed a marked reduction in macroporosity, water retention, and plant-available water, whereas fine-texture soils showed smaller changes in water availability but reduced aeration associated with macropore reduction. These results indicate higher physical quality vulnerability of coarse-textured soils following forest-to-pasture conversion. Full article

Mind–body practices integrating movement and meditation, such as Tai Chi and Aikido, have been proposed to influence the gut–brain axis through combined physiological and psychological pathways. However, evidence regarding their association with gut microbiota composition remains limited. This study explored gut microbiota diversity and taxonomic profiles in regular practitioners of Tai Chi and Aikido across different levels of meditation depth. Forty-two adults practicing Tai Chi or Aikido provided fecal samples for 16S rRNA sequencing, and meditation depth was assessed using the Meditation Depth Questionnaire (MEDEQ). Alpha diversity did not differ significantly between groups, although a descriptive trend toward higher diversity with increasing meditation depth was observed. Beta-diversity analyses suggested compositional differences associated with meditation level (ANOSIM R = 0.191, p = 0.035), along with an exploratory interaction signal between practice type and meditation depth (ANOSIM R = 0.296, p = 0.001). Taxonomic profiling highlighted distinct microbial patterns associated with both practice type and meditation depth. Short-chain fatty acid-associated genera, including Faecalibacterium and Roseburia, were relatively more abundant in Aikido practitioners with higher meditation scores, whereas Tai Chi practitioners showed higher relative abundances of Prevotella and Collinsella. Overall, these findings indicate that meditative movement practices are associated with distinct gut microbiota compositional patterns within this cohort. Given the exploratory and cross-sectional design, these results should be interpreted as hypothesis-generating. Future longitudinal studies incorporating functional and clinical outcomes are needed to clarify underlying mechanisms. Full article

Background: Pulmonary infarction (PI) is the result of the occlusion of distal pulmonary arteries resulting in damage to downstream lung areas that become ischemic, hemorrhagic, or necrotic, and it is often a complication of an underlying condition such as pulmonary embolism (PE). Since in most of cases it is located peripherally, lung ultrasound (LUS) can be a good evaluation tool. The typical radiological features of PI are well-known; however, there are limited data on its sonographic characteristics and its evolution. Methods: The aim of this study is to evaluate, using LUS, a convenience sample of patients with acute PE with computed tomography (CT) consolidation findings consistent with PI. Patients’ clinical characteristics were collected and LUS findings at baseline and their short-term progression was assessed. LUS was performed within 72 h of PE diagnosis (T0) and repeated after one (T1) and four weeks (T2). Each procedure started with a focused examination of the areas of lesions based on CT findings, followed by an exploration of the other posterior and lateral lung fields. The convex probe was used for initial evaluation integrating LUS evaluation with the linear one was employed for smaller and more superficial lesions and when appropriate. Color Doppler mode was added to study vascularization. Results: From June to October 2023, 14 consecutive patients were enrolled at the Respiratory Unit of the University Hospital of Pisa. The main population characteristics included the absence of respiratory failure and prognostic high-risk PE (100%), the absence of significant comorbidities (79%), and the presence of typical symptoms, such as chest pain (57%) and dyspnea (50%). The average number of consolidations per patient was 1.4 ± 0.6. Follow-up LUS showed the disappearance of some consolidations and some morphological changes in the remaining lesions: the presence of hypoechoic consolidation with a central hyperechoic area (“bubbly consolidation”) was more typical at T1 while the presence of a small pleural effusion often persisted both at T1 and T2. A decrease in wedge/triangular-shaped consolidations was observed (82% at T0, 67% at T1, 24% at T2), as was an increase in elongated shapes, representing a residual pleural thickening over time (9% at T0, 13% at T1, 44% at T2). A reduction in size was also observed by comparing the mean diameter, long axis, and short axis measurements of each consolidation at the three different studied time points: the average of the short axes and the median of the mean diameters showed a statistically significant reduction after four weeks. Additionally, a correlation between lesion size and pleuritic pain was described, although it did not achieve statistical significance. Conclusions: Patients’ clinical characteristics and ultrasound features are consistent with previous studies studying PI at PE diagnosis. Most consolidations detected by LUS change over time regarding size and form, but a minority of them do not differ. LUS is a safe and non-invasive exam that could help to improve patients’ clinical approach in emergency rooms as well as medical and pulmonology settings, clinically contextualized for cases of chest pain and dyspnea. Future studies could expand the morphological study of PI. Full article