Search Results (646)

The rapid progress of generative speech synthesis and voice-cloning technologies has enabled the creation of highly natural synthetic voices that pose a serious threat to telecommunication security. While most prior studies evaluate human ability to detect audio deepfakes using high-quality, studio-grade recordings, little is known about how real-world telecommunication channels affect perceptual detection. This study investigates the influence of three transmission scenarios—GSM (AMR-NB), VoLTE (AMR-WB), and VoIP with packet-loss modeling—on the human ability to distinguish natural speech from AI-generated speech. A custom speech corpus was developed, consisting of natural recordings from nine speakers and corresponding synthetic utterances generated using a state-of-the-art voice cloning system (ElevenLabs). All samples were processed through simulated telecommunication channels using real codec implementations. A listening test with 95 participants was conducted, involving binary classification (human vs. synthetic) and confidence ratings. Results show an overall detection accuracy of 54.8%, confirming that humans are poorly equipped to identify synthetic speech. Surprisingly, the highest accuracy was achieved for the narrowband GSM channel (63.7%), while VoLTE yielded the lowest performance (44.0%). The findings suggest that restricted bandwidth may emphasize prosodic irregularities typical of generative models, whereas high-quality channels mask synthetic artifacts, increasing susceptibility to voice spoofing. The results highlight the necessity of deploying additional security mechanisms in telecommunication systems relying on voice identity verification. Full article

Feline Infectious Peritonitis (FIP) has been transformed from a fatal disease to a treatable condition following the introduction of GS-441524, a nucleoside analogue targeting feline coronavirus replication. However, the widespread use of unregulated compounded formulations and the absence of validated analytical tools for therapeutic drug monitoring (TDM) represent critical gaps in clinical FIP management. This study describes the development and full ICH M10-compliant validation of a high-throughput LC-MS/MS method for the quantification of GS-441524 in feline serum, incorporating an automated protein precipitation protocol and a PBS-BSA surrogate matrix in accordance with 3Rs principles. The method met all acceptance criteria across validated parameters, including linearity (0.1–50 µg/mL), accuracy (bias within ±12.5%), precision (CV ≤ 10.9%), selectivity, extraction recovery (87.5–107.9%), and stability under clinically relevant storage conditions. Matrix equivalence between PBS-BSA and authentic feline serum was confirmed, enabling routine calibration without animal-derived materials. The validated method was applied to clinical TDM in cats undergoing GS-441524 treatment for FIP, providing preliminary evidence of inter-individual pharmacokinetic variability. The compounded formulations administered to the TDM cohort were independently verified by LC-MS/MS, confirming drug content within ±15% of labelled claims and excluding pharmaceutical quality as a confounding factor in the interpretation of serum drug concentrations. Full article

Chagas disease, caused by Trypanosoma cruzi (T. cruzi), includes clinically relevant intestinal inflammation; however, the mechanisms associated with tissue injury remain incompletely understood. ArtinM is an immunomodulatory lectin with known effects on innate and adaptive immunity, although its intestinal role during acute T. cruzi infection remains unclear. This study investigated whether ArtinM modulates the intestinal inflammatory response during acute experimental T. cruzi infection. In vivo, BALB/c mice were allocated to Saline control, T. cruzi + Saline, and T. cruzi + ArtinM groups. Intestinal inflammatory infiltrate and tissue concentrations of TNF-α, IFN-γ, IL-12p40, and IL-10 were quantified. Acute infection markedly increased TNF-α, IFN-γ, IL-12p40, and inflammatory infiltrate, whereas ArtinM significantly attenuated these responses. TNF-α, IFN-γ, and IL-12p40 remained associated with group after adjustment for infiltrate, whereas IL-10 reached statistical significance only in the adjusted model and was therefore interpreted cautiously. In parallel, an exploratory analysis of a public murine intestinal scRNA-seq dataset (GSE319934; GSM9529706 and GSM9529707), derived from a chronic infection setting, was performed to provide pathway-level context for inflammatory mediators assessed in vivo. This transcriptomic analysis indicated that related inflammatory, innate immune, chemotactic, and adhesion-associated genes were detectable in intestinal single-cell data from T. cruzi infection. However, because this dataset was not temporally matched to the acute model, it was not interpreted as a phase-matched comparator, mechanistic validation, or temporal extension of the experimental findings. Together, the results support that ArtinM treatment is associated with attenuation of acute intestinal inflammatory outcomes in experimental T. cruzi infection. Because local intestinal parasite burden was not measured, these findings should be interpreted as evidence of inflammatory modulation rather than as direct evidence of local antiparasitic activity. The public scRNA-seq analysis provides only exploratory contextual information for related inflammatory pathways. Full article

This study investigates the thermochemical reaction behaviour of scheelite–millscale aluminothermy for direct tungsten alloying in steel production. Experimental mixtures of aluminium, millscale, and scheelite concentrate were simulated using gas–slag–metal (g-s-m) equilibrium calculations in FactSage 8.3 at 2200 °C, and compared with previously reported experimental results. The simulations reproduced metal yields accurately with 0.901 to 0.940 correlation coefficients and predicted tungsten levels consistent with measured steel compositions. However, significant discrepancies were observed in predicted silicon levels, with simulations overestimating steel %Si by up to 3.5%, despite negligible gas-phase losses. Oxygen partial pressure calculations indicate that the Fe/FeO reaction equilibrium controls process reduction conditions. Backcalculation of activity coefficients revealed that FactSage minimisation routines understated silicon activity coefficient values. SiO₂ mass transfer may play a role in low %Si in steel, but this is not clear due to differences in expected mass transfer regimes in aluminothermy under ASR and SHS conditions. Overall, the simulations demonstrate adequate predictive capability for alloying trends and metal yields while highlighting limitations in predicting silicon partitioning. These findings confirm the utility of thermochemical simulation for designing aluminothermic feed mixtures, reducing the number of experiments needed to optimise the aluminothermic feed mixture ratios. Full article

To identify quality markers (Q-markers) for daodi authenticity evaluation of Lycium barbarum L., a comprehensive strategy integrating appearance trait analysis, multi-component quantification, and chemometrics was developed. Forty-five sample batches were collected from four major producing areas in China, namely Ningxia (NX), Gansu (GS), Qinghai (QH), and Inner Mongolia (NM). Appearance traits (50-fruit weight, moisture, and color) and the contents of polysaccharide, total sugar, betaine, zeaxanthin dipalmitate, and 27 small-molecule compounds, including flavonoids and phenolics, were determined using UV–vis spectrophotometry, HPLC-CAD, and UPLC-MS/MS. Pearson correlation analysis revealed a significant negative association between polysaccharide and total sugar (r = −0.344, p < 0.05), suggesting a possible allocation shift between the two carbohydrate fractions, while zeaxanthin dipalmitate strongly correlated with redness (r = 0.609, p < 0.01). Principal component analysis identified total sugar, polysaccharide, scopoletin, and scopolin as key discriminatory variables. AHP-CRITIC combined weighting highlighted polysaccharide (weight 0.195) and zeaxanthin dipalmitate (weight 0.157) as candidate core Q-markers. Top-ranked comprehensive scores predominantly belonged to samples from NX and GS, chemically supporting the traditional daodi authenticity. This dual-dimensional “efficacy–trait” framework provides a robust, traceable basis for origin authentication and quality standard improvement of L. barbarum. Full article

Robust Arithmetic Word Problem (AWP) solving is important for applying mathematical reasoning systems in educational scenarios, where problem statements may contain changed numerical values, paraphrased descriptions, or irrelevant distracting information. Although Large Language Models (LLMs) have shown strong potential in solving AWPs, their reasoning processes may still be sensitive to surface-form variations and perturbation-induced noise. To address this issue, this paper proposes a Scene-Aware Neuro-Symbolic solver designed to improve the robustness of AWP solving under perturbations. The proposed method extends the existing scene-aware framework by introducing perturbation-oriented mechanisms at the scene, relation, and symbolic-solving levels. A Chain-of-Scene (CoS) prompting strategy first generates candidate scenes, after which goal-guided filtering retains target-related and bridge scenes while removing distractor-induced scenes. The retained scenes are then processed by the Scene-Aware Syntax-Semantics ($S^2$) method to extract explicit and implicit relations, and relation consistency checking is applied to remove locally plausible but globally irrelevant relations. Finally, the symbolic solver performs iterative equation-based reasoning over the filtered relation sets, with fallback recovery activated when standard solving does not produce a target-compatible answer. Experiments on AGG, MAWPS, and GSM8K show an average accuracy of 92.8% on clean datasets. On GSM-Perturb and AWP-Perturb, the solver achieves perturbed accuracies of 80.8% and 87.5%, with robustness drops of 8.3% and 6.8%, respectively. Ablation results show that scene filtering and relation consistency checking are the main contributors to reducing perturbation-induced errors. These findings suggest that combining LLM-based scene understanding with symbolic relation reasoning is a promising direction for improving the robustness and interpretability of AWP solvers in the evaluated perturbation settings. Full article

Microfinance institutions (MFIs) are vital in promoting financial inclusion for underserved populations. However, these institutions face growing security threats, including sophisticated burglary tactics like underground tunneling. In the Philippines, notable incidents, such as the “Termite Gang” heist in Marikina City and a mall robbery in Ozamiz, highlight the limitations of conventional security systems in addressing subterranean intrusions. This study addresses the gap in existing security technologies by developing a real-time detection system that integrates a vibration sensor, a Global System for Mobile Communications (GSM) module for sending real-time SMS alerts, an audible alarm, and a solar-powered backup system for continuous operation. The system was simulated in the electrical technology laboratory to enhance classroom learning. The system’s core is an Arduino Uno microcontroller that processes inputs from the SW-420 vibration sensor, activating alarms and triggering SMS notifications via the SIM900A module when it detects unusual vibrations. Simulations A, B, and C were conducted to evaluate the system’s response time, with results showing a progressive reduction in detection time from five seconds to one second, indicating improved calibration and system efficiency. These findings also support the existing literature on user interaction with vibration alerts, demonstrating high accuracy in interpreting haptic notifications and the cognitive trade-offs involved. The proposed solution offers a proactive, energy-resilient, and cost-effective security system specifically designed to address underground burglary attempts. It applies to MFIs, pawnshops, and other high-risk financial environments. Future research should explore the application of machine learning for adaptive threat detection, expand the system’s scalability, and integrate mobile applications to enable user customization and enhance alert management. Full article

Excess zinc (Zn) has an important effect on seed yield and quality, as well as the possibility of subsequent processing. In this study, we investigated the effect of excess Zn concentrations (50, 100, and 150 mg kg⁻¹) in the substrate on the seed yield and selected biochemical parameters of white mustard (Sinapis alba L.) seeds. The following parameters were investigated: seed yield, individual lipid classes were analyzed using high-performance thin-layer chromatography (HPTLC); gas–liquid chromatography with mass-selective detection (GS-MS) was used to analyze the fatty acid (FA) profile; and trace elements were detected using inductively coupled plasma-mass spectrometry (ICP-MS). It was found that Zn at concentrations of 100 and 150 mg kg⁻¹ caused a decrease in the number of S. alba pods, and the 1000-seed weight also decreased at a Zn concentration of 150 mg kg⁻¹. The Zn concentration in seeds from plants grown on contaminated substrate was higher than the control values and government standard thresholds. Zn at concentrations of 100 and 150 mg kg⁻¹ caused a slight increase in the content of triacylglycerols (up to 1.35–1.36% dry weight) and the total content of unsaturated FAs along with a decrease in saturated and total FAs. Zn at a concentration of 50 mg kg⁻¹ stimulated an increase in the total FA content. A high erucic acid content in S. alba seeds was observed in all studied variants. A mathematical model was used to evaluate the physicochemical properties of biofuels. The parameters relating to the FA composition of seeds of plants grown on a Zn-contaminated substrate did not deteriorate, and in the case of the kinematic viscosity coefficient, they were even improved compared to the control plants and complied with American (ASTM D6751) and European (EN 14214) standards for biofuels. The obtained data indicate a negative effect of high Zn concentrations on seed yield. Changes in the FA composition of S. alba seeds may reduce their value for the production of edible oils due to the high content of erucic acid, but they could be used to produce technical oils (biofuel). Full article

Valorization of agro-industrial waste into functional materials is fundamental to the circular economy, especially for addressing the persistent contamination by anionic azo dyes in textile wastewater. This study evaluates guava seeds modified with hexadecyltrimethylammonium bromide (GS-M) as low-cost biosorbents for the removal of Direct Blue 71 (DB71), comparing their performance with that of natural seeds (GS-N) in batch systems and fixed-bed columns. Characterization by infrared spectroscopy (FTIR) and electron microscopy (SEM-EDS) confirmed successful surfactant immobilization, thereby creating a cationic surface with strong electrostatic affinity for anionic dye molecules. Batch experiments showed that GS-M achieved 98% DB71 removal within 120 min, whereas GS-N reached only 58% after 300 min. For GS-M, both pseudo-first-order and pseudo-second-order models fit the kinetic data well, consistent with concurrent electrostatic and hydrophobic interactions; GS-N was best described by the Elovich model, indicating rate limitation by electrostatic repulsion. GS-M maintained removal efficiency above 84% across pH 3–9, whereas GS-N was effective under acidic conditions. Langmuir maximum adsorption capacity (Q_o) values for GS-M were 6.02 mg/g at pH 4 and 7.87 mg/g at pH 8, a 1.5- to 2.2-fold increase over GS-N under matched conditions. Three adsorption–desorption cycles retained ~49% of the initial GS-M capacity, supporting a short-cycle reuse profile rather than indefinite multi-cycle operation. Fixed-bed column performance was highly sensitive to the hydraulic loading rate (v_c), with breakthrough times increasing nearly eightfold as v_c decreased. The Bed Depth Service Time (BDST), Thomas, and Yoon–Nelson models described the dynamic data consistently, yielding a maximum dynamic capacity of 165.6 mg/L under optimal conditions and providing a quantitative basis for scale-up. These results establish surfactant-modified guava seeds as a low-cost, pH-resilient biosorbent system aligned with circular-economy principles for the sustainable remediation of textile wastewater. Full article

The increased deployment of photovoltaic (PV) technology has led to an increased demand for grid-balancing capacity owing to growing short-term variability and forecast uncertainty. Simultaneously, higher PV penetration can lead to daytime energy market oversupply, pushing day-ahead prices toward zero and undermining PV revenues. Against this backdrop, this study investigated a market participation paradigm in which PV power plants supply reserve power themselves while actively absorbing their own uncertainty, rather than merely relying on balancing the services provided by external resources. We propose a risk-aware framework that classifies solar irradiance prediction errors into four risk categories using GPV-GSM numerical weather forecast data, translating the inferred risk level into practical bidding rules for balancing market participation. We adopted a hierarchical classification pipeline consisting of sign determination (stage 1, under- vs. overprediction), followed by degree determination (Stages 2 and 3), implemented with a multi-layer perceptron. To enhance class separability and reduce features, we introduced a stage-wise area under the curve (AUC)-based feature selection and compared AUC-selected and all-features settings under identical training conditions. The proposed strategies substantially reduce shortage events compared with directly using the original predictions as bids, although they increase surplus energy. The AUC-based model achieves comparable imbalance evaluation results, indicating that the selected features are sufficient for practical bidding support. Full article

The fashion industry generates significant environmental impact through overproduction and the short life cycles of clothing products. Modular design, which allows multiple wearing configurations from a single product through removable and interchangeable components, represents a promising sustainable alternative but lacks standardized methodologies that integrate digital innovation with physical production. This study develops and validates an integrated methodology for creating versatile modular clothing products. Based on the three-dimensional framework of modularity (structure, function, and system), four categories of modular elements were defined and organized in a CLO3D (Version 2026.0.202; CLO Virtual Fashion Inc., Seoul, Republic of Korea) digital library. Combinatorial calculation generated 360 theoretical variants, of which 24 were selected as structurally and aesthetically feasible. A physical prototype consisting of a vest, four modules and three sleeve variants was made from 100% linen fabrics (290 and 180 gsm) with snap fastener tape and natural shell buttons. The digital–physical visual comparison confirmed high concordance for proportions and alignment of the modules, with limitations in simulating the fabric drape. The 24 validated configurations demonstrate that a single modular clothing product can function as an effective wardrobe multiplier. The study confirms that integrating the CLO3D digital library with physical prototyping from sustainable materials constitutes a viable and replicable methodology in the fashion industry. Full article

Mathematical problem transformation is a teaching-and-learning strategy that extends conceptual understanding and problem-solving ability by expressing the same concept across diverse situations. It has recently attracted attention in artificial intelligence as a tool for data augmentation, difficulty control, and model evaluation. However, existing approaches struggle to jointly represent and control how core mathematical elements—such as operational structure, quantitative relations, and conditions—are preserved or modified. This limitation is particularly evident in natural-language problems, where intertwined components make it difficult to perform targeted partial transformations or verify structural validity. To address these challenges, we propose the Symbolic Math Problem Representation Framework (SyMPRep), which represents the relationships among sentences, conditions, questions, quantities, units, and operations in a symbolic structure. It classifies free-form instructions into predefined categories and decomposes problems into constituent elements, enabling transformation over an explicit abstraction structure. This allows problem transformation to be treated as a controllable, traceable, and recoverable structural operation rather than surface rewriting. Experiments on GSM8K and Math500 show that SyMPRep achieves stable alignment and recoverability, and confirm that the main challenge lies in structural control rather than surface fluency. Ablation results highlight the importance of symbolic schema and show that different metrics capture distinct aspects of transformation quality. In downstream applications, answer-invariant transformations yield modest improvements on easier problems, while human evaluation indicates that the generated problems are coherent and suitable for educational use. These findings suggest that SyMPRep serves as a representation-driven interface for controllable structural transformation. Full article

In this paper, we investigate the propagation properties of a partially coherent Gaussian Schell-model (GSM) beam by effective beam parameters in oceanic turbulence. We provide detailed analytical derivations based on the extended Huygens–Fresnel integral and the cross-spectral density function. It is found that the angle-of-arrival fluctuation, spreading, and wander of the partially coherent GSM beam decrease with increasing source coherence parameter and turbulent kinetic energy dissipation rate, and with decreasing temperature fluctuations and mean-square temperature dissipation rate. At 200 m propagation distance, the relative mean-squared width under salinity-dominated conditions (ω = −2) is approximately 0.02% larger than that under temperature-dominated conditions (ω = −5), indicating that salinity fluctuations cause more obvious beam spreading. Full article

Environmental monitoring systems are often constrained by high cost, limited portability, restricted pollutant coverage, and dependence on fixed infrastructure, which can limit their suitability for distributed real-time sensing. This study presents FRESH, an autonomous Internet of Things (IoT)-based platform for multi-parameter environmental monitoring and short-term forecasting. The system integrates sensors for air quality, thermal conditions, light, acoustics, and weather, together with GSM-based remote data transmission, onboard data logging, and hybrid battery–solar power management. FRESH was deployed across multiple indoor and outdoor locations in Coventry and at the University of Warwick, UK, and operated over a 10-month period to assess practical performance under varied environmental conditions. In addition to continuous environmental sensing, machine learning models were developed to predict short-term changes in selected environmental variables. Across the tested models, the best predictive performance was obtained for several key parameters, including particulate matter (R² = 0.93), volatile organic compounds (R² = 0.92), and ozone (R² = 0.98). The results suggest that FRESH has potential to support portable, multi-parameter environmental monitoring with integrated short-horizon forecasting, providing a basis for further development of distributed sensing and localised early-warning applications. Full article

Parameter-efficient fine-tuning (PEFT) enables large language models to adapt to downstream tasks with low computational cost. As a representative high-rank PEFT method, MoRA (High-Rank Updating for Parameter-Efficient Fine-Tuning) improves update expressiveness through a compression–transformation–decompression reparameterization mechanism. However, its bottleneck subspace is still modeled using a freely learned linear transformation. In addition, grouped compression may project information from different original directions into shared bottleneck coordinates. This may reduce subspace separability and lead to inefficient utilization of the effective update space. To address this limitation, we propose GL-log-MoRA, which introduces a learnable general linear transformation into the MoRA bottleneck subspace and applies log-determinant regularization to encourage a more balanced spectral structure. In this way, the proposed method improves directional coordination and subspace expressiveness without imposing hard structural constraints or causing noticeable memory overhead. We evaluate GL-log-MoRA on five benchmarks: LogiQA, Financial PhraseBank, GSM8K, FinQA, and HotpotQA. The results show that GL-log-MoRA achieves the best performance on these downstream tasks and yields small but consistent improvements over MoRA under the same parameter budget. Compared with MoRA, GL-log-MoRA improves LogiQA from 42.50% to 45.45% and Financial PhraseBank from 81.60% to 83.02%. It also improves GSM8K from 63.1% to 64.6%, FinQA from 10.02% to 10.23%, and HotpotQA from 70.6% to 70.8%. Meanwhile, the average empirical effective-rank indicator increases from 1.05 to 2.80. Peak GPU memory changes only slightly, from 18.21 GB to 18.28 GB. Full article