Search Results (60,159)

Weakly supervised video anomaly detection (WSVAD) aims to identify unusual events using only video-level labels. However, current methods face several key challenges, including ineffective modelling of complex temporal dependencies, indistinct feature boundaries between visually similar normal and abnormal events, and high false alarm rates caused by an inability to distinguish salient events from complex background noise. This paper proposes a novel method that systematically enhances feature representation and discrimination to address these challenges. The proposed method first builds robust temporal representations by employing a hierarchical multi-scale temporal encoder and a position-aware global relation network to capture both local and long-range dependencies. The core of this method is the dual-stream attention-enhanced memory network, which achieves precise discrimination by learning distinct normal and abnormal patterns via dual memory banks, while utilising bidirectional spatial attention to mitigate background noise and focus on salient events before memory querying. The models underwent a comprehensive evaluation utilising solely RGB features on two demanding public datasets, UCF-Crime and XD-Violence. The experimental findings indicate that the proposed method attains state-of-the-art performance, achieving 87.43% AUC on UCF-Crime and 85.51% AP on XD-Violence. This result demonstrates that the proposed “attention-guided prototype matching” paradigm effectively resolves the aforementioned challenges, enabling robust and precise anomaly detection. Full article

Background: Spinal anesthesia is a widely used technique for pain control in surgical procedures, requiring effective drug distribution within the cerebrospinal fluid (CSF) for optimal outcomes. The distribution is influenced by injection parameters such as needle diameter and injection speed, which, if not optimized, can reduce efficacy or cause side effects. This study investigates how these parameters affect drug distribution in the CSF using computational fluid dynamics (CFD). Material Methods: An anatomically accurate three-dimensional model of the CSF space was created using MRI data. Simulations were performed using three needle tips (22 G, 25 G, 27 G) and different injection rates at the L4–L5 vertebral level. The model included physiological CSF oscillations from cardiac and respiratory cycles. Drug dispersion was analyzed in terms of spatial distribution and concentration changes over time. Results: The findings obtained show that the combination of a large-gauge needle (22G) and high injection speed provides wider distribution within the CSF and more effective transport to the cranial regions. On the other hand, with a small-gauge needle (27G) and low injection speed, the drug remained more localized, and access to the upper spinal regions was limited. Additional parameters such as injection duration, direction, and flush applications were also observed to significantly affect distribution. Conclusions: CFD modeling reveals that injection parameters significantly affect drug dispersion patterns in spinal anesthesia. Optimizing these parameters may improve therapeutic outcomes and reduce complications. The model provides a foundation for developing personalized intrathecal injection protocols. Full article

Ionospheric scintillation represents a disturbance phenomenon induced by irregular electron density variations, predominantly occurring in equatorial, auroral, and polar regions, thereby posing significant threats to Global Navigation Satellite Systems (GNSS) performance. Polar regions in particular confront distinctive challenges, including the sparse deployment of dedicated ionospheric scintillation monitoring receiver (ISMR) equipment, the limited availability of strong scintillation samples, severely imbalanced training datasets, and the insufficient sensitivity of conventional Deep Neural Networks (DNNs) to intense scintillation events. To address these challenges, this study proposes a modeling framework that integrates residual neural networks (ResNet) with the Synthetic Minority Over-sampling Technique for Regression with Gaussian Noise (SMOGN). The proposed model incorporates multi-source disturbance features to accurately estimate phase scintillation indices (σ_φ) in polar regions. The methodology was implemented and validated across multiple polar observation stations in Canada. Shapley Additive Explanations (SHAP) interpretability analysis reveals that the rate of total electron content index (ROTI) features contribute up to 64.09% of the predictive weight. The experimental results demonstrate a substantial performance enhancement compared with conventional DNN models, with root mean square error (RMSE) values ranging from 0.0078 to 0.038 for daytime samples in 2024, and an average coefficient of determination (R²) consistently exceeding 0.89. The coefficient of determination for the Pseudo-Random Noise (PRN) path estimation results can reach 0.91. The model has good estimation results at different latitudes and is able to accurately capture the distribution characteristics of the local strong scintillation structures and their evolution patterns. Full article

Background/Objectives: Vertical growth pattern significantly influences facial aesthetics and treatment choices. Lateral cephalograms are routinely used for the evaluation of vertical jaw relationships in orthodontic diagnosis. The aim of this study was to evaluate the performance of deep learning algorithms in classifying cephalometric radiographs according to vertical skeletal growth patterns without the need for anatomical landmark identification. Methods: This study was carried out on lateral cephalometric radiographs of 1050 patients. Cephalometric radiographs were divided into 3 subgroups based on FMA, SN-GoGn, and Cant of Occlusal Plane angles. Six deep learning models (ResNet101, DenseNet 201, EfficientNet B0, EfficientNet V2 B0, ConvNetBase, and a hybrid model) were employed for the classification of the dataset. The performances of the well-known deep learning models and the hybrid model were compared for accuracy, precision, F1-Score, mean absolute error, Cohen’s Kappa, and Grad-CAM metrics. Results: The highest accuracy rates were achieved by the Hybrid Model with 86.67% for FMA groups, 87.29% for SN-GoGn groups, and 82.71% for Cant of Occlusal Plane groups. The lowest accuracy rates were achieved by ConvNet with 79.58% for FMA groups, 65% for SN-GoGn, and 70.21% for Cant of Occlusal Plane groups. Conclusions: The six deep learning algorithms employed demonstrated classification success rates ranging from 65% to 87.29%. The highest classification accuracy was observed in the FMA angle, while the lowest accuracy was recorded for the Cant of the Occlusal Plane angle. The proposed DL algorithms showed potential for direct skeletal orthodontic diagnosis without the need for cephalometric landmark detection steps. Full article

For aquaculture systems, maximizing feed efficiency is a major challenge since it directly affects growth rates and economic sustainability. Feed is one of the largest costs in aquaculture, and feed waste is a significant environmental issue that requires effective management strategies. This paper suggests a novel approach for optimal fish feeding in pond aquaculture systems that integrates vision language models (VLMs), optical flow, and advanced image processing techniques to enhance feed management strategies. The system allows for the precise assessment of fish needs in connection to their feeding habits by integrating real-time data on biomass estimates and water quality conditions. By combining these data sources, the system makes informed decisions about when to activate automated feeders, optimizing feed distribution and cutting waste. A case study was conducted at a profit-driven tilapia farm where the system had been operational for over half a year. The results indicate significant improvements in feed conversion ratios (FCR) and a 28% reduction in feed waste. Our study found that, under controlled conditions, an average of 135 kg of feed was saved daily, resulting in a cost savings of approximately $1800 over the course of the study. The VLM-based fish feeding behavior recognition system proved effective in recognizing a range of feeding behaviors within a complex dataset in a series of tests conducted in a controlled pond aquaculture setting, with an F1-score of 0.95, accuracy of 92%, precision of 0.90, and recall of 0.85. Because it offers a scalable framework for enhancing aquaculture resource use and promoting sustainable practices, this study has significant implications. Our study demonstrates how combining language models and image processing could transform feeding practices, ultimately improving aquaculture’s environmental stewardship and profitability. Full article

This study analyses the impacts of winter storms on beach response, as well as the subsequent recovery during spring and summer, at a dissipative sandy beach in Brandon Bay, Ireland. Shoreline dynamics were assessed through the integration of field data from five survey campaigns conducted between October 2021 and November 2022 with a 1D Xbeach (version 1.23) numerical model. Cross-sectional profiles were measured at seven locations, revealing pronounced erosion during winter, followed by recovery in calmer seasons, especially in the lower beach zone. The model effectively simulated short-term storm-induced morphological changes, demonstrating that rates of shoreline retreat and profile alteration are higher in the eastern bay, where wave energy is greater. Most morphological changes occurred between the low and high astronomical tide marks, characterized by upper beach erosion and lower beach accretion. Models were subsequently employed to examine future climate scenarios, including sea level rise and increased storm intensity. The projections indicated an exponential increase in erosion rates, correlated with higher storm wave heights and frequencies. These results highlight the dynamic response of dissipative beaches to extreme events and reinforce the necessity for adaptive coastal management strategies to address the escalating risks posed by climate change. Full article

This study presents a three-dimensional numerical analysis of natural convection during the postharvest storage of corn and wheat in a galvanized steel silo with a conical roof and floor, measuring 3 m in radius and 18.7 m in height, located in the Bajío region of Mexico. Simulations were carried out specifically for December, a period characterized by cold ambient temperatures (10–20 °C) and comparatively lower solar radiation than in warmer months, yet still sufficient to induce significant heating of the silo’s metallic surfaces. The governing conservation equations of mass, momentum, energy, and species were solved using the finite volume method under the Boussinesq approximation. The model included grain–air sorption equilibrium via sorption isotherms, as well as metabolic heat generation: for wheat, a constant respiration rate was assumed due to limited biochemical data, whereas for corn, respiration heat was modeled as a function of grain temperature and moisture, thereby more realistically representing metabolic activity. The results, obtained for December storage conditions, reveal distinct thermal and hygroscopic responses between the two grains. Corn, with higher thermal diffusivity, developed a central thermal core reaching 32 °C, whereas wheat, with lower diffusivity, retained heat in the upper region, peaking at 29 °C. Radial temperature profiles showed progressive transitions: the silo core exhibited a delayed response relative to ambient temperature fluctuations, reflecting the insulating effect of grain. In contrast, grain at 1 m from the wall displayed intermediate amplitudes. In contrast, zones adjacent to the wall reached 40–41 °C during solar exposure. In comparison, shaded regions exhibited minimum temperatures close to 15 °C, confirming that wall heating is governed primarily by solar radiation and metal conductivity. Axial gradients further emphasized critical zones, as roof-adjacent grain heated rapidly to 38–40 °C during midday before cooling sharply at night. Relative humidity levels exceeded 70% along roof and wall surfaces, leading to condensation risks, while core moisture remained stable (~14.0% for corn and ~13.9% for wheat). Despite the cold ambient temperatures typical of December, neither temperature nor relative humidity remained within recommended safe storage ranges (10–15 °C; 65–75%). These findings demonstrate that external climatic conditions and solar radiation, even at reduced levels in December, dominate the thermal and hygroscopic behavior of the silo, independent of grain type. The identification of unstable zones near the roof and walls underscores the need for passive conservation strategies, such as grain redistribution and selective ventilation, to mitigate fungal proliferation and storage losses under non-aerated conditions. Full article

The stochastic model in Precise Point Positioning (PPP) defines the statistical properties of observations and the dynamic behavior of parameters. An inaccurate stochastic model can degrade positioning accuracy, ambiguity resolution, and other aspects of performance. However, due to the influence of multiple factors, the stochastic model in PPP cannot be precisely predetermined, necessitating the development of an Adaptive Stochastic Model (ASM) based on Variance Component Estimation (VCE). While the benefits of ASMs for PPP float solutions are well documented, their contributions to other performance aspects remain insufficiently explored. This paper presents a comprehensive assessment of an ASM’s impact on PPP. First, the implementation of an ASM using VCE is described in detail. Then, experimental results demonstrate that the ASM effectively captures observational conditions through the estimated variance component factors. It enhances both PPP float and fixed solutions when the predefined stochastic model is inadequate, improves cycle-slip detection by tightening the stochastic model (reducing the missed detection rate from 19% to 8%), and accelerates both direct reconvergence and re-initialization after data gaps, with reconvergence times improved by 18% and 55%, respectively. Full article

Objectives: Virtual reality (VR) has emerged as a promising tool for assessing anxiety-related disorders through immersive exposure and physiological monitoring. This study aimed to evaluate whether multimodal data, including heart rate variability (HRV), skin conductance response (SCR), and self-reported anxiety, collected during VR exposure could classify patients with panic disorder and agoraphobia using machine learning models. Methods: Seventy-six participants (38 patients with panic disorder and agoraphobia, 38 healthy controls) completed 295 total VR exposure sessions. Each session involved two road and two supermarket scenarios designed to induce anxiety. Inside the sessions, self-reported anxiety was measured along with physiological signals recorded by photoplethysmography and SCR sensors. HRV measures of heart rate, standard deviation of normal-to-normal intervals, and low-frequency to high-frequency ratio were extracted along with SCR peak frequency and average amplitude. These features were analyzed using Gaussian Naïve Bayes (GNB), k-Nearest Neighbors (k-NN), Logistic Ridge Regression (LRR), C-Support Vector Machine (SVC), Random Forest (RF), and Stochastic Gradient Boosting (SGB) classifiers. Results: The best model achieved an accuracy of 0.83. Most models showed specificity and precision ≥0.80, while sensitivity varied across models, with several reaching ≥0.82. Performance was stable across major hyperparameters, VR-stimulus settings, and medication status. The patients reported higher subjective anxiety but exhibited blunted physiological responses, particularly in SCR amplitude. Self-reported anxiety demonstrated higher feature importance scores compared to other physiological properties. Conclusion: VR exposure with self-reported anxiety and physiological measures may serve as a feasible diagnostic aid for panic disorder and agoraphobia. Further refinement is needed to improve sensitivity and clinical applicability. Full article

One of the key components of the software quality model is reliability. Its importance has grown with the increasing use and reuse of open-source components in software development. Software reliability growth models are commonly employed to address this aspect by predicting future failure rates and estimating the number of remaining defects throughout the development process. This paper investigates two software reliability growth models derived from the Verhulst model, with a particular focus on a structural property known as Hausdorff saturation. We provide analytical estimates for this characteristic and propose it as an additional criterion for model selection. The models are evaluated using four open-source datasets, where the Hausdorff saturation metric supports the conclusions drawn from standard goodness-of-fit measures. Furthermore, we introduce an interactive software reliability assessment tool that integrates with GitHub, enabling expert users to analyze real-time issue-tracking data from open-source repositories. The tool facilitates model comparison and enhances practical applicability. Overall, the proposed approach contributes to more robust reliability assessment by combining theoretical insights with actionable diagnostics. Full article

This study proposes a task planning approach for a distributed constellation dedicated to space target monitoring, grounded in an adaptive genetic algorithm. The approach is designed to address challenges such as the growing number of space targets and the complex constraints inherent in space target monitoring activities. After reviewing the research progress of distributed satellite task planning and adaptive genetic algorithms, a distributed task model featuring master-slave satellites was developed. This model integrates multi-constraint modeling and aims to optimize key performance indicators: task yield rate, task completion rate, resource utilization rate, and load balancing. To enhance the approach, the contract net algorithm is fused with the adaptive genetic algorithm: Firstly, in the tendering phase, centralized tendering is adopted to reduce communication overhead; Secondly, in the bidding phase, improved genetic mechanisms (e.g., dynamic reverse adjustment of crossover and mutation probabilities) and a dynamic population strategy are employed to generate task allocation schemes; Thirdly, in the bid evaluation and winning phase, differentiated strategies are applied to non-repetitive and repetitive tasks. Simulation validation shows that this approach can complete 80% of space target monitoring tasks, balance satellite loads effectively, and manage space target catalogs efficiently. Full article

Phenol is a refractory organic pollutant that is difficult to degrade in wastewater treatment, and efficiently and stably degrading phenol presents a significant challenge. In this study, iron-doped humic acid-based nitrogen–carbon materials were prepared to activate peroxymonosulfate (PMS) for the degradation of phenol. The Fe-N-C/PMS system achieved a phenol degradation rate of 99.71%, which follows a first-order kinetic model, with the reaction rate constant of 0.1419 min⁻¹. The phenol degradation rate remained above 92% in inorganic anions (Cl⁻, SO₄²⁻, HCO₃⁻) and humic acid and the system maintained a 100% phenol removal rate over a wide pH range (3–9). The iron in the catalyst predominantly exists in the forms of Fe⁰ and Fe₃C, and Fe⁰, Fe²⁺/Fe³⁺ are the main active sites that promote PMS activation during the reaction. Additionally, Fe-N-C has a large specific surface area (1041.36 m²/g). Quenching experiments and electron spin resonance (ESR) spectroscopy detected the active free radicals in the Fe-N-C/PMS system: SO₄^•−, •OH, O₂^•−, and ¹O₂. The mechanism for phenol degradation was discussed, involving radical pathways (SO₄^•−, •OH, O₂^•−) and the non-radical pathway (¹O₂), in the Fe-N-C/PMS system activated by Fe⁰, Fe²⁺/Fe³⁺, sp² hybridized carbon, C-O/C-N, C=O, and graphitic nitrogen active sites. This study provides new insights into the synthesis of efficient carbon-based catalysts for phenol degradation and water remediation. Full article

Obstructive sleep apnea–hypopnea syndrome (OSAHS) is a prevalent sleep disorder strongly linked to increased cardiovascular and metabolic risk. While prior studies have explored snore-based analysis for OSAHS, they have largely focused on either detection or classification in isolation. Here, we present a two-stage framework that integrates precise snoring event detection with deep learning-based classification. In the first stage, we develop an Adaptive Multi-Feature Fusion Endpoint Detection algorithm (AMFF-ED), which leverages short-time energy, spectral entropy, zero-crossing rate, and spectral centroid to accurately isolate snore segments following spectral subtraction noise reduction. Through adaptive statistical thresholding, joint decision-making, and post-processing, our method achieves a segmentation accuracy of 96.4%. Building upon this, we construct a balanced dataset comprising 6830 normal and 6814 OSAHS-related snore samples, which are transformed into Mel spectrograms and input into ERBG-Net—a hybrid deep neural network combining ECA-enhanced ResNet18 with bidirectional GRUs. This architecture captures both spectral patterns and temporal dynamics of snoring sounds. The experimental results demonstrate a classification accuracy of 95.84% and an F1 score of 94.82% on the test set, highlighting the model’s robust performance and its potential as a foundation for automated, at-home OSAHS screening. Full article

Prepartum vaccination of dairy cows against newborn calf diarrhea protects calves during the first weeks of life via the colostrum. Vaccination may also induce non-specific effects (NSEs) beyond antibody production, altering the disease susceptibility and productivity of the vaccinated mother. This retrospective study analyzed herd records and on-site survey data from 73,378 dairy cows on 20 German farms using linear mixed-effects models and random forest algorithms. Management practices and milk yield showed stronger associations with outcomes than vaccination. However, the cows vaccinated with non-live vaccines had increased odds of retained placenta and metritis (OR: 1.5–1.7), as well as endometritis (OR: 3–6), and were 20–24% less likely to conceive than non-vaccinated cows. Among non-live vaccinated cows, those vaccinated 2.5–4 weeks before calving had an 8% higher non-return rate compared to those vaccinated 6–8 weeks prior. Multiparous cows receiving live vaccine components were 1.9 times more likely to conceive, compared to non-live vaccinated multiparous cows. These findings suggest potential NSE of prepartum vaccination on uterine health and fertility. However, this study’s retrospective design limits causal interpretation, and the benefits in calves may outweigh possible adverse effects. Further research should clarify the mechanisms and optimize vaccine timing and composition. Full article

In the polyurethane industry, catalytically generated carbodiimides can modify the properties of isocyanate and, thus, the resulting foams. In this work, a kinetic reaction study was carried out to investigate the formation of a simple, bifunctional carbodiimide from a widely used polyurethane raw material: 4,4′-methylenediphenyl diisocyanate (MDI). The experimental section outlines a catalytic process, using a 3-methyl-1-phenyl-2-phospholene-1-oxide (MPPO) catalyst in ortho-dichlorobenzene (ODCB) solvent, to model industrial circumstances. The reaction produces carbon dioxide, which was observed using gas volumetry at between 50 and 80 °C to obtain kinetic data. A detailed regression analysis with linear and novel nonlinear fits showed that the initial stage of the reaction is second-order, and the temperature dependence of the rate constant is $k(T)=(3.4\pm 3.8)\mathrm{‧}{10}^6\cdot e^{{\displaystyle \frac{-7192\pm 389}{T}}}$. However, the other isocyanate group of MDI reacts with new isocyanate groups and the reaction deviates from the second-order due to oligomer (polycarbodiimide) formation and other side reactions. A linearized Arrhenius equation was used to determine the activation energy of the reaction, which was E_a = 60.4 ± 3.0 kJ mol⁻¹ at the applied temperature range, differing by only 4.6 kJ mol⁻¹ from a monoisocyanate-based carbodiimide. In addition to experimental results, computationally derived thermochemical data (from simplified DFT and IRC calculations) were applied in transition state theory (TST) for a comprehensive prediction of rate constants and Arrhenius parameters. As a result, it was found that the activation energy of the carbodiimide bond formation reaction from theoretical and experimental results was independent of the number and position of isocyanate groups, which is consistent with the principle of equal reactivity of functional groups. Full article