Search Results (848)

Anomaly detection in industrial imaging is critical for ensuring quality and reliability in automated manufacturing processes. While recently several methods have been reported in the literature that have demonstrated impressive detection performance on standard benchmarks, they necessarily rely on computationally intensive CNN architectures and post-processing techniques, necessitating access to high-end GPU hardware and limiting practical deployment in resource-constrained settings. In this study, we introduce a novel anomaly detection framework that leverages feature maps from a lightweight convolutional neural network (CNN) backbone, MobileNetV2, and cascaded detection to achieve notable accuracy as well as computational efficiency. The core of our method consists of two main components. First is a PCA-based anomaly detection module that specifically exploits near-zero variance features. Contrary to traditional PCA methods, which tend to focus on the high-variance directions that encapsulate the dominant patterns in normal data, our approach demonstrates that the lower variance directions (which are typically ignored) form an approximate null space where normal samples project near zero. However, the anomalous samples, due to their inherent deviations from the norm, lead to projections with significantly higher magnitudes in this space. This insight not only enhances sensitivity to true anomalies but also reduces computational complexity by eliminating the need for operations such as matrix inversion or the calculation of Mahalanobis distances for correlated features otherwise needed when normal behavior is modeled as Gaussian distribution. Second, our framework consists of a cascaded multi-stage decision process. Instead of combining features across layers, we treat the local features extracted from each layer as independent stages within a cascade. This cascading mechanism not only simplifies the computations at each stage by quickly eliminating clear cases but also progressively refines the anomaly decision, leading to enhanced overall accuracy. Experimental evaluations on MVTec and VisA benchmark datasets demonstrate that our proposed approach achieves superior anomaly detection performance (99.4% and 91.7% AUROC respectively) while maintaining a lower computational overhead compared to other methods. This framework provides a compelling solution for practical anomaly detection challenges in diverse application domains where competitive accuracy is needed at the expense of minimal hardware resources. Full article

Potatoes are one of the world’s most widely cultivated crops, but their yield is coming under mounting pressure from early blight, a fungal disease caused by Alternaria solani. Early detection and accurate identification are key to effective disease management and yield protection. This paper introduces a novel deep learning framework called ZeroShot CNN, which integrates convolutional neural networks (CNNs) and ZeroShot Learning (ZSL) for the efficient classification of seen and unseen disease classes. The model utilizes convolutional layers for feature extraction and employs semantic embedding techniques to identify previously untrained classes. Implemented on the Kaggle potato disease dataset, ZeroShot CNN achieved 98.50% accuracy for seen categories and 99.91% accuracy for unseen categories, outperforming conventional methods. The hybrid approach demonstrated superior generalization, providing a scalable, real-time solution for detecting agricultural diseases. The success of this solution validates the potential in harnessing deep learning and ZeroShot inference to transform plant pathology and crop protection practices. Full article

This study addresses the understanding of fungal diversity and their bioremediation roles in an integrated aquaculture wastewater bioremediation system, an area less explored compared to bacteria, viruses, and protozoa. Despite the rapid advancement and affordability of molecular tools, insights into fungal communities remain vague, and interpreting environmental studies in an ecologically meaningful manner continues to pose challenges. To bridge this knowledge gap, we developed an integrated aquaculture wastewater bioremediation system, incorporating photosynthetic bacteria, and utilizing internal transcribed spacer (ITS) sequencing to analyze fungal community composition. Our findings indicate that the fungal community in aquaculture wastewater is predominantly composed of the phyla Ascomycota and Chytridiomycota, with dominant genera including Aspergillus, Hortea, and Ciliphora. FUNGuild, a user-friendly trait and character database operating at the genus level, facilitated the ecological interpretation of fungal functional groups. The analysis revealed significant negative correlations between nutrient levels (COD_mn, NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, and PO₄⁻³-P) and specific fungal functional groups, including epiphytes, animal pathogens, dung saprotrophs, plant pathogens, and ectomycorrhizal fungi. The removal rate for the COD_mn, NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, and PO₄⁻³-P were 71.42, 91.37, 88.80, 87.20, and 91.72% respectively. This study highlights the potential role of fungal communities in bioremediation processes and provides a framework for further ecological interpretation in aquaculture wastewater treatment systems. Full article

The increasing need for materials that are both lightweight and strong in the aerospace and automotive sectors has driven the extensive use of composite sandwich structures. This study examines the impact response of honeycomb sandwich composites fabricated using the vacuum-assisted resin transfer molding (VARTM) technique. Two configurations were analyzed, namely carbon–honeycomb–carbon (CHC) and carbon–Kevlar–honeycomb–Kevlar–carbon (CKHKC), to assess the effect of Kevlar reinforcement on impact resistance. Charpy impact testing was conducted to evaluate energy absorption, revealing that CKHKC composites exhibited significantly superior impact resistance compared to CHC composites. The CKHKC composite achieved an average impact strength of 70.501 KJ/m², which is approximately 73.8% higher than the 40.570 KJ/m² recorded for CHC. This improvement is attributed to Kevlar’s superior toughness and energy dissipation capabilities. A comparative assessment of impact energy absorption further highlights the advantages of hybrid Kevlar–carbon fiber composites, making them highly suitable for applications requiring enhanced impact performance. These findings provide valuable insights into the design and optimization of high-performance honeycomb sandwich structures for impact-critical environments. Full article

The present study provides a detailed compilation and analysis of the bulk density and magnetic susceptibility of the rocks from the northwest Himalayas, Pakistan. The area is tectonically extremely complex and comprises sedimentary, metamorphic, and igneous rocks. These rocks range in age from Early Proterozoic to Recent. During the fieldwork, 476 rock samples were collected for density measurements and 410 for magnetic susceptibility measurements from the major rock units exposed in the study area. The measured physical parameters reveal a significant difference in the density and susceptibility of the rocks present in the investigated area. The sedimentary rock units belonging to the Indian Plate show the lowest mean values for bulk density, followed by metasedimentary rocks, Early Proterozoic rocks, igneous and metaigneous rock units of the Indian Plate, Indus Suture Melange Zone, and Kohistan Island Arc rocks, respectively. The magnetic susceptibility of sedimentary rock units of the Indian Plate has the lowest mean values, followed by metasedimentary rocks of the Indian Plate, igneous and metaigneous rock units of the Indian Plate, Early Proterozoic rocks of the Indian Plate, Kohistan Island Arc rocks, and Indus Suture Melange Zone. In brief, the sedimentary rocks of the Indian Plate have the lowest bulk density and magnetic susceptibility values, whereas the Kohistan Island Arc rocks have the highest values. Overall, the bulk density and magnetic susceptibility of rock units in the study area follow those predicted for different types of rocks. These measurements can be used to develop possible potential field models of the northwest Himalayas to better understand the tectonics of the ongoing continental-to-continental collision, as well as for many other geological analyses. Full article

The performance of the UAVs while executing various mission profiles greatly depends on the selection of planning algorithms. Reinforcement learning (RL) algorithms can effectively be utilized for robot path planning. Due to random action selection in case of action ties, the traditional Q-learning algorithm and its other variants face the issues of slow convergence and suboptimal path planning in high-dimensional navigational environments. To solve these problems, we propose an improved deep Q-network (DQN), incorporating an efficient tie-breaking mechanism, prioritized experience replay (PER), and L2-regularization. The adopted tie-breaking mechanism improves the action selection and ultimately helps in generating an optimal trajectory for the UAV in a 3D cluttered environment. To improve the convergence speed of the traditional Q-algorithm, prioritized experience replay is used, which learns from experiences with high temporal difference (TD) error and avoids uniform sampling of stored transitions during training. This also allows the prioritization of high-reward experiences (e.g., reaching a goal), which helps the agent to rediscover these valuable states and improve learning. Moreover, L2-regularization is adopted that encourages smaller weights for more stable and smoother Q-values to reduce the erratic action selections and promote smoother UAV flight paths. Finally, the performance of the proposed method is presented and thoroughly compared against the traditional DQN, demonstrating its superior effectiveness. Full article

The demand for rare earth elements (REEs) is continuously increasing due to the important roles they play in low-carbon and green energy technologies. Unfortunately, the global REE reserves are limited and concentrated in only a few countries, so the reprocessing of alternative resources like tailings is of critical importance. This study investigated carrier magnetic separation using coarse magnetite particles as a carrier to recover finely ground monazite from tailings. The monazite and carrier surfaces were modified by sodium oleate (NaOL) to improve the hydrophobic interactions between them. The results of zeta potential and contact angle measurements implied the selective adsorption of NaOL onto the surfaces of the monazite and magnetite particles. Although their hydrophobicity increased, heterogenous agglomeration between them was not substantial. To improve heterogenous agglomeration, emulsified kerosene was utilized as a bridging liquid, resulting in more extensive attachment of fine monazite particles onto the surfaces of carrier particles and a dramatic improvement in monazite recovery by magnetic separation—from 0% (without carrier) to 70% (with carrier). A rougher–scavenger–cleaner carrier magnetic separation can produce REE concentrates with a total rare earth oxide (TREO) recovery of 80% and a grade of 9%, increased from 3.4%, which can be further increased to 23.2% after separating REEs and the carrier. Full article

Biomimetic design for engineering applications may suggest the optimal performance of engineering devices. In this work the passive/pure pitching characteristics of a hydrofoil are investigated experimentally with and without a pair of biomimetic fin strips placed symmetrically on the two sides of the foil leading edge. The work is performed in a recirculating water channel at low Reynolds numbers (Re) with a range of 1300 ≤ Re ≤ 3200. Using high-speed videography and Particle Image Velocimetry (PIV), the pitching characteristics and wakes are visualized. Passive pitching characteristics, i.e., the pitching amplitude and pitching frequency of the hydrofoils, are investigated based on their trailing edge movement. Significant improvement in both pitching frequency and amplitudes are observed for the foil with fin strips compared to the baseline simple foil. Comparing the pitching characteristics of the two foils, it is observed that the hydrofoil with biomimetic fin strips exhibits 25% and 21% higher pitching amplitude and pitching frequency, respectively, compared to that of the baseline at comparable Reynolds numbers. The initiation of pitching for the finned foil is also observed at comparatively low Reynolds numbers. The wake is also studied using time mean and fluctuating velocity profiles obtained using PIV. Full article

Background: Cardiac arrest remains a significant public health challenge with variable mortality trends across different demographics and regions, affecting healthcare planning and intervention strategies. We conducted this study to analyze cardiac arrest-related mortality trends from 1999 to 2023 and predict future trends up to 2035. Methods: This study analyzed data from 1999 to 2023, focusing on cardiac arrest as the primary cause of death (ICD-10: I46). Age-adjusted mortality rates (AAMRs) were standardized according to the 2000 U.S. Census. Joinpoint regression was utilized to calculate annual percentage change (APC), and an ARIMA model with Python 3.10 was used for mortality predictions. Results: A total of 365,608 cardiac arrest-related deaths were recorded in the USA from 1999 to 2023. There was a sharp decline in mortality rate until 2001 (APC: −10.35, p < 0.05), followed by a slowed decline until 2013 (APC: −2.91, p < 0.05), and then a gradual uptrend. Males exhibited a higher AAMR (5.8, 95% CI: 5.8–5.9) compared to females (4.2, 95% CI: 4.1–4.2). African Americans had the highest AAMR (8.9, 95% CI: 8.9–9), followed by Caucasians (4.8, 95% CI: 4.8–4.9) and American Indians (3.5, 95% CI: 3.3–3.7). The South region of the US had the highest AAMR, followed by the Northeast, Midwest, and West. Alabama exhibited the highest AAMR, followed by Nevada and Hawaii. Predictive analysis suggests a potential stable slow downtrend in mortality rates by 2035 (AAMR: 4.28, 95% CI: −1.8–10.4). Conclusions: The observed trends and future predictions underscore the importance of targeted public health interventions and healthcare planning to address cardiac arrest mortality. Full article

As the world accelerates toward a sustainable future with electric vehicles (EVs), smartphone applications have become an indispensable tool for drivers. These applications, developed by both EV manufacturers and third-party developers, offer functionalities such as remote vehicle control, charging station location, and route planning. However, they also have access to sensitive information, making them potential targets for cyber threats. This paper presents a comprehensive survey of the cybersecurity vulnerabilities, weaknesses, and permissions in these applications. We categorize 20 applications into two groups: those developed by EV manufacturers and those by third parties, and conduct a comparative analysis of their functionalities by performing static and dynamic analysis. Our findings reveal major security flaws such as poor authentication, broken encryption, and insecure communication, among others. The paper also discusses the implications of these vulnerabilities and the risks they pose to users. Furthermore, we analyze 10 permissions and 12 functionalities that are not present in official EV applications and mostly present in third-party apps, leading users to rely on poorly built third-party applications, thereby increasing their attack surface. To address these issues, we propose defensive measures which include 10 CWE AND OWASP top 10 defenses to enhance the security of these applications, ensuring a safe and secure transition to EVs. Full article

Recurrent catastrophic inverter failures significantly undermine the reliability and economic viability of utility-scale photovoltaic (PV) power plants. This paper presents a comprehensive investigation of severe inverter destruction incidents at the Kopli Solar Power Plant, Estonia, by integrating controlled laboratory simulations with extensive field monitoring. Initially, detailed laboratory experiments were conducted to replicate critical DC-side short-circuit scenarios, particularly focusing on negative DC input terminal faults. The results consistently showed these faults rapidly escalating into multi-phase short-circuits and sustained ground-fault arcs due to inadequate internal protection mechanisms, semiconductor breakdown, and delayed relay response. Subsequently, extensive field-based waveform analyses of multiple inverter failure events captured identical fault signatures, thereby conclusively validating laboratory-identified failure mechanisms. Critical vulnerabilities were explicitly identified, including insufficient isolation relay responsiveness, inadequate semiconductor transient ratings, and ineffective internal insulation leading to prolonged arc conditions. Based on the validated findings, the paper proposes targeted inverter design enhancements—particularly advanced DC-side protective schemes, rapid fault-isolation mechanisms, and improved internal insulation practices. Additionally, robust operational and monitoring guidelines are recommended for industry-wide adoption to proactively mitigate future inverter failures. The presented integrated methodological framework and actionable recommendations significantly contribute toward enhancing inverter reliability standards and operational stability within grid-connected photovoltaic installations. Full article

Heart attack is a leading cause of mortality, necessitating timely and precise diagnosis to improve patient outcomes. However, timely diagnosis remains a challenge due to the complex and nonlinear relationships between clinical indicators. Machine learning (ML) and deep learning (DL) models have the potential to predict cardiac conditions by identifying complex patterns within data, but their “black-box” nature restricts interpretability, making it challenging for healthcare professionals to comprehend the reasoning behind predictions. This lack of interpretability limits their clinical trust and adoption. The proposed approach addresses this limitation by integrating predictive modeling with Explainable AI (XAI) to ensure both accuracy and transparency in clinical decision-making. The proposed study enhances heart attack prediction using the University of California, Irvine (UCI) dataset, which includes various heart analysis parameters collected through electrocardiogram (ECG) sensors, blood pressure monitors, and biochemical analyzers. Due to class imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) was applied to enhance the representation of the minority class. After preprocessing, various ML algorithms were employed, among which Artificial Neural Networks (ANN) achieved the highest performance with 96.1% accuracy, 95.7% recall, and 95.7% F1-score. To enhance the interpretability of ANN, two XAI techniques, specifically SHapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME), were utilized. This study incrementally benchmarks SMOTE, ANN, and XAI techniques such as SHAP and LIME on standardized cardiac datasets, emphasizing clinical interpretability and providing a reproducible framework for practical healthcare implementation. These techniques enable healthcare practitioners to understand the model’s decisions, identify key predictive features, and enhance clinical judgment. By bridging the gap between AI-driven performance and practical medical implementation, this work contributes to making heart attack prediction both highly accurate and interpretable, facilitating its adoption in real-world clinical settings. Full article

This article addresses the non-trivial problem of classifying thoracic diseases in chest X-ray (CXR) images. A single CXR image may exhibit multiple diseases, making this a multi-label classification problem. Additionally, the inherent class imbalance makes the task even more challenging as some diseases occur more frequently than others. Our methodology is based on transfer learning aiming to fine-tune a pretrained DenseNet121 model using CXR images from the NIH Chest X-ray14 dataset. Training from scratch would require a large-scale dataset containing millions of images, which is not available in the public domain for this multi-label classification task. To address class imbalance problem, we propose a rank-based loss derived from the Zero-bounded Log-sum-exp and Pairwise Rank-based (ZLPR) loss, which we refer to as focal ZLPR (FZLPR). In designing FZLPR, we draw inspiration from the focal loss where the objective is to emphasize hard-to-classify examples (instances of rare diseases) during training compared to well-classified ones. We achieve this by incorporating a “temperature” parameter to scale the label scores predicted by the model during training in the original ZLPR loss function. Experimental results on the NIH Chest X-ray14 dataset demonstrate that FZLPR loss outperforms other loss functions including binary cross entropy (BCE) and focal loss. Moreover, by using test-time augmentations, our model trained using FZLPR loss achieves an average AUC of 80.96% which is competitive with existing approaches. Full article

Understanding the complex interplay among altitudinal gradients, tree species diversity, structural attributes, and soil carbon (C) is critical for effective coniferous forest management and climate change mitigation. This study addresses a knowledge gap by investigating the effects of altitudinal gradient on coniferous tree diversity, biomass, carbon stock, regeneration, and soil organic carbon storage (SOCs) in the understudied temperate forests of the Hindu-Kush Kalam Valley. Using 120 sample plots 20 × 20 m (400 m²) each via a field inventory approach across five altitudinal gradients [E1 (2000–2200 m)–E5 (2801–3000 m)], we comprehensively analyzed tree structure, composition, and SOCs. A total of four coniferous tree species and 2172 individuals were investigated for this study. Our findings reveal that elevation indirectly influences species diversity, SOCs, and forest regeneration. Notably, tree height has a positive relationship with altitudinal gradients, while tree carbon stock exhibits an inverse relationship. Forest disturbance was high in the middle elevation gradients E2–E4, with high deforestation rate at E1 and E2. Cedrus deodara, the dominant species, showed the highest deforestation rate at lower elevations (R² = 0.72; p < 0.05) and regeneration ability (R² = 0.77; p < 0.05), which declined with increasing elevation. Middle elevations had the highest litter carbon stock and SOCs values emphasizing the critical role of elevation gradients in carbon sink and species distribution. The regeneration status and number of trees per ha in Kalam Valley forests showed a significant decline with increasing elevation (p < 0.05), with Cedrus deodara recording the highest regeneration rate at E1 and Abies pindrow the lowest at E5. The PCA revealed that altitudinal gradients factor dominate variability via PCA1, while the Shannon and Simpson Indices drives PCA2, highlighting ecological diversity’s independent role in shaping distinct yet complementary vegetative and ecological perspectives. This study reveals how altitudinal gradients shape forest structure and carbon sequestration, offering critical insights for biodiversity conservation and climate-resilient forest management. Full article

A continuously variable transmission can improve the energy efficiency of actuators with rotary output by providing an optimum transmission ratio. A continuously variable transmission based on circumferentially arranged disks (CAD CVT) is a new type of CVT that is highly beneficial for applications requiring large torques, like heavy road transport. However, its major drawback is that its efficiency drops in the low torque region. To overcome this problem, the current paper proposes an improved mechanical design in which the force on traction disks is changed according to the instantaneous torque requirement, thus resulting in improved efficiency in low torque regions. Furthermore, a hydraulic-actuation-based control system has been designed to ensure the optimum control of the improved mechanical design. The improved mechanical design of the CAD CVT is named CAD CVT-II, which is highly beneficial for variable torque applications such as road transport and wind turbines. Full article