Search Results (7,063)

The accurate prediction of herb–target interactions is essential for the modernization of traditional Chinese medicine (TCM) and the advancement of drug discovery. Nonetheless, the inherent complexity of herbal compositions and diversity of molecular targets render experimental validation both time-consuming and labor-intensive. We propose a graph neural network model, MAMGN-HTI, which integrates metapaths with attention mechanisms. A heterogeneous graph consisting of herbs, efficacies, ingredients, and targets is constructed, where semantic metapaths capture latent relationships among nodes. An attention mechanism is employed to dynamically assign weights, thereby emphasizing the most informative metapaths. In addition, ResGCN and DenseGCN architectures are combined with cross-layer skip connections to improve feature propagation and enable effective feature reuse. Experiments show that MAMGN-HTI outperforms several state-of-the-art methods across multiple metrics, exhibiting superior accuracy, robustness, and generalizability in HTI prediction and candidate drug screening. Validation against literature and databases further confirms the model’s predictive reliability. The model also successfully identified herbs with potential therapeutic effects for hyperthyroidism, including Vinegar-processed Bupleuri Radix (Cu Chaihu), Prunellae Spica (Xiakucao), and Processed Cyperi Rhizoma (Zhi Xiangfu). MAMGN-HTI provides a reliable computational framework and theoretical foundation for applying TCM in hyperthyroidism treatment, providing mechanistic insights while improving research efficiency and resource utilization. Full article

High-precision semantic segmentation of remote sensing imagery is crucial in geospatial analysis. It plays an immeasurable role in fields such as urban governance, environmental monitoring, and natural resource management. However, when confronted with complex objects (such as winding roads and dispersed buildings), existing semantic segmentation methods still suffer from inadequate target recognition capabilities and multi-scale representation issues. This paper proposes a neural network model, LMVMamba (LoRA Multi-scale Vision Mamba), for semantic segmentation of remote sensing images. This model integrates the advantages of convolutional neural networks (CNNs), Transformers, and state-space models (Mamba) with a multi-scale feature fusion strategy. It simultaneously captures global contextual information and fine-grained local features. Specifically, in the encoder stage, the ResT Transformer serves as the backbone network, employing a LoRA fine-tuning strategy to effectively enhance model accuracy by training only the introduced low-rank matrix pairs. The extracted features are then passed to the decoder, where a U-shaped Mamba decoder is designed. In this stage, a Multi-Scale Post-processing Block (MPB) is introduced, consisting of depthwise separable convolutions and residual concatenation. This block effectively extracts multi-scale features and enhances local detail extraction after the VSS block. Additionally, a Local Enhancement and Fusion Attention Module (LAS) is added at the end of each decoder block. LAS integrates the SimAM attention mechanism, further enhancing the model’s multi-scale feature fusion capability and local detail segmentation capability. Through extensive comparative experiments, it was found that LMVMamba achieves superior performance on the OpenEarthMap dataset (mIoU 52.3%, OA 69.8%, mF1: 68.0%) and LoveDA (mIoU 67.9%, OA 80.3%, mF1: 80.5%) datasets. Ablation experiments validated the effectiveness of each module. The final results indicate that this model is highly suitable for high-precision land-cover classification tasks in remote sensing imagery. LMVMamba provides an effective solution for precise semantic segmentation of high-resolution remote sensing imagery. Full article

Magnetohydrodynamic (MHD) flow and heat transfer in porous media are central to many engineering applications, including heat exchangers, MHD generators, and polymer processing. This study examines the boundary layer flow and thermal behavior of an electrically conducting viscous fluid over a porous stretching tube. The model accounts for nonlinear thermal radiation, internal heat generation/absorption, and Darcy–Forchheimer drag to capture porous medium resistance. Similarity transformations reduce the governing equations to a system of coupled nonlinear ordinary differential equations, which are solved numerically using the BVP4C technique with Response Surface Methodology (RSM) and sensitivity analysis. The effects of dimensionless parameters magnetic field strength (M), Reynolds number (Re), Darcy–Forchheimer parameter (Df), Brinkman number (Br), Prandtl number (Pr), nonlinear radiation parameter (Rd), wall-to-ambient temperature ratio (rw), and heat source/sink parameter (Q) are investigated. Results show that increasing M, Df, and Q suppresses velocity and enhances temperature due to Lorentz and porous drag effects. Higher Re raises pressure but reduces near-wall velocity, while rw, Rd, and internal heating intensify thermal layers. The entropy generation analysis highlights the competing roles of viscous, magnetic, and thermal irreversibility, while the Bejan number trends distinctly indicate which mechanism dominates under different parameter conditions. The RSM findings highlight that rw and Rd consistently reduce the Nusselt number (Nu), lowering thermal efficiency. These results provide practical guidance for optimizing energy efficiency and thermal management in MHD and porous media-based systems.: Full article

Particle accelerators are powerful tools in fundamental research, medicine, and industry that provide high-energy beams that can be used to study matter and to enable advanced applications. The state-of-the-art particle accelerators are fundamentally constructed from superconducting radio-frequency (SRF) cavities, which act as resonant structures for the acceleration of charged particles. The performance of such cavities is governed by inherent superconducting material properties such as the transition temperature, critical fields, penetration depth, and other related parameters and material quality. For the last few decades, bulk niobium has been the preferred material for SRF cavities, enabling accelerating gradients on the order of ~50 MV/m; however, its intrinsic limitations, high cost, and complicated manufacturing have motivated the search for alternative strategies. Among these, sputter-deposited superconducting thin films offer a promising route to address these challenges by reducing costs, improving thermal stability, and providing access to numerous high-T_c superconductors. This review focuses on progress in sputtered superconducting materials for SRF applications, in particular Nb, NbN, NbTiN, Nb₃Sn, Nb₃Al, V₃Si, Mo–Re, and MgB₂. We review how deposition process parameters such as deposition pressure, substrate temperature, substrate bias, duty cycle, and reactive gas flow influence film microstructure, stoichiometry, and superconducting properties, and link these to RF performance. High-energy deposition techniques, such as HiPIMS, have enabled the deposition of dense Nb and nitride films with high transition temperatures and low surface resistance. In contrast, sputtering of Nb₃Sn offers tunable stoichiometry when compared to vapour diffusion. Relatively new material systems, such as Nb₃Al, V₃Si, Mo-Re, and MgB₂, are just a few of the possibilities offered, but challenges with impurity control, interface engineering, and cavity-scale uniformity will remain. We believe that future progress will depend upon energetic sputtering, multilayer architectures, and systematic demonstrations at the cavity scale. Full article

Sketch re-identification (Re-ID) aims to retrieve pedestrian photographs in the gallery dataset by a query sketch image drawn by professionals, which is crucial for criminal investigations and missing person searches in the field of public security. The main challenge of this task lies in bridging the significant modality gap between sketches and photos while extracting discriminative modality-invariant features. However, information asymmetry between sketches and RGB photographs, particularly the differences in color information, severely interferes with cross-modal matching processes. To address this challenge, we propose a novel network architecture that integrates multi-channel augmentation with hierarchically consistent disentanglement learning. Specifically, a multi-channel augmentation module is developed to mitigate the interference of color bias in cross-modal matching. Furthermore, a modality-disentangled prototype(MDP) module is introduced to decompose pedestrian representations at the feature level into modality-invariant structural prototypes and modality-specific appearance prototypes. Additionally, a cross-layer decoupling consistency constraint is designed to ensure the semantic coherence of disentangled prototypes across different network layers and to improve the stability of the whole decoupling process. Extensive experimental results on two public datasets demonstrate the superiority of our proposed approach over state-of-the-art methods. Full article

Due to its efficiency, advanced oxidation processes (AOP), such as photo-Fenton, have become an alternative for removing emerging contaminants like paracetamol. The objective of this work was to perform a life cycle assessment (LCA) according to ISO 14040/44 for a heterogeneous photo-Fenton process catalyzed by Cu/Fe-pillared clays (PILC) for the removal of paracetamol from water. The study covered catalyst synthesis and four treatment scenarios, with inventories built from experimental data and ecoinvent datasets; treatment time was 120 min per functional unit. Environmental impacts for catalyst synthesis were quantified with ReCiPe 2016 (midpoint), while toxicity-related impacts of the degradation stage were assessed with USEtox™ (human carcinogenic toxicity, human non-carcinogenic toxicity, and freshwater ecotoxicity). Catalyst synthesis dominated most midpoint categories, the global warming potential for 1 g of Cu/Fe-PILC was 10.98 kg CO₂ eq. Toxicity results for S4 (photo-Fenton Cu/Fe PILC) showed very low values: 9.73 × 10⁻¹² CTUh for human carcinogenic and 1.29 × 10⁻¹³ CTUh for human non-carcinogenic. Freshwater ecotoxicity ranged from 5.70 × 10⁻⁴ PAF·m³·day at pH 2.7 (≥60 min) to 1.67 × 10⁻⁴ PAF·m³·day at pH 5.8 (120 min). Overall, optimizing pH and reaction time, are key levers to improve the environmental profile of AOP employing Cu/Fe-PILC catalysts. Full article

Static mixers are commonly used for process intensification in a wide range of industrial applications. For the design and selection of a static mixer, an accurate prediction of the hydraulic performance, particularly the pressure drop, is essential. This experimental study examines the pressure drop for turbulent single-phase and gas–liquid two-phase flows through a Komax triple-action static mixer placed on a horizontal pipeline. New values of friction factor and z-factor are reported for fully turbulent liquid single-phase flow (11,700 ≤ Re_L ≤ 18,700). For two-phase flow, the pressure drop for stratified and intermittent flows (0.07 m/s ≤ U_L ≤ 0.28 m/s and 0.46 m/s ≤ U_G ≤ 3.05 m/s) is modeled using the Lockhart–Martinelli approach, with a coefficient, C, correlated to the homogenous void fraction. Conversely, the analysis of power dissipation reveals a dependence on both liquid and gas superficial velocities. For conditions corresponding to intermittent flow upstream of the mixer, flow visualization revealed the emergence of a swirling flow in the Komax static mixer. It is interesting to note that an increase in slug frequency leads to an increase, followed by stabilization of the pressure drop. The results offer valuable insights for improving the design and optimization of Komax static mixers operating under single-phase and two-phase flow conditions. In particular, the reported correlations can serve as practical tools for predicting hydraulic losses during the design and scale-up. Moreover, the observed influence of the slug frequency on the pressure drop provides guidance for selecting operating conditions that minimize energy consumption while ensuring efficient mixing. Full article

Intelligent sensing systems are increasingly used in wearable devices, enabling advanced tasks across various application domains including robotics and human–machine interaction. Ensuring these systems are energy autonomous is highly demanded, despite strict constraints on power, memory and processing resources. To meet these requirements, embedded neural networks must be optimized to achieve a balance between accuracy and efficiency. This paper presents an integrated approach that combines Hardware-Aware Neural Architecture Search (HW-NAS) with optimization techniques—weight reshaping, quantization, and their combination—to develop efficient classifiers for a multisensory glove. HW-NAS automatically derives 1D-CNN models tailored to the NUCLEO-F401RE board, while the additional optimization further reduces model size, memory usage, and latency. Across three datasets, the optimized models not only improve classification accuracy but also deliver an average reduction of 75% in inference time, 69% in flash memory, and more than 45% in RAM compared to NAS-only baselines. These results highlight the effectiveness of integrating NAS with optimization techniques, paving the way towards energy-autonomous wearable systems. Full article

As native bovine whey (WHEY) poses environmental concerns as a high-water-content by-product, this trial aimed at assessing the effectiveness of a thermal–mechanical microparticulation coupled with a fermentative process to concentrate it into a high-protein soft dairy cream. Compared to native whey, in microparticulated (MPW) and fermented (FMPW) matrices, there was a significant increase in proteins (from 0.7 to 8.8%) and lipids (from 0.3 to 1.3%), and a more brilliant yellowness colour. A factorial discriminant analysis (FDA) showed that FMPW had a higher content of saturated fatty acid (SFA) and some specific polyunsaturated fatty acid (PUFA) n-6, and also identified C14:0, C18:1, C18:1 t-11, C18:2 n-6, and C18:3 n-6 as informative biomarkers of microparticulation and fermentative treatments. The SDS-PAGE indicated no effects on the protein profile but indicated its rearrangement into high molecular weight aggregates. Z-sizer and transmission electron microscopy analyses confirmed a different supramolecular structure corresponding to a higher variability and greater incidence of very large molecular aggregates, suggesting that MPW could be accounted as a colloidal matrix that may have similar ball-bearing lubrication properties. Microparticulation of whey could facilitate its circularity into the dairy supply chain through its re-generation from a waste into a high-value fat replacer for dairy-based food production. Full article

Fibre-reinforced composites are facing new challenges in the context particular in sustainability and recyclability. Vitrimers could be useful as new matrices to support the increase in sustainability. Due to their high strength, which is comparable to that of thermosets often used in composites, and their covalent adaptive networks, which make them reshapeable for scaled-up manufacturing and recycling purposes, they are very useful. Orthogonal cutting is used for precise reshaping and functional integration into carbon fibre reinforced plastics. Vitrimers could improve processing results at the cutting edge as well as surface quality thanks to their self-healing properties compared to brittle matrices, as well as enabling the recycling of formed chips and scrap. This study showcases the manufacturing of a carbon fibre-reinforced vitrimer using 4-aminophenyl disulfide as a hardener, with vacuum-assisted resin infusion. The temperature of chip formation and the cutting parameters are then shown for different fibre orientations, cutting widths and speeds. The observed cutting forces are lower (less than 140 N) and more irregular for fibre orientations 45°/135°, increasing with cutting depth, and fluctuating periodically during machining. Despite varying cutting speeds, the forces remain relatively constant in range between 85 N and 175 N for 0°/90° fibre orientation and 50 N and 120 N for 45°/135° fibre orientation, with no significant tool wear observed and lower-damage depth and overhanging fibres observed for 0°/90° fibre orientation. Damage observation of the cutting tool shows promising results, with lower abrasion observed compared to thermoset matrices. Microscopic images of the broached surface also show good quality, which could be improved by self-healing of the matrix at higher temperatures. Temperature measurements of chip formation using a high-speed camera show a high temperature gradient as cutting speeds increase, but the temperature only ever exceeds 180 °C at cutting speeds of 150 m/min, ensuring reprocessability since this is below the degradation temperature. Therefore, orthogonal cutting of vitrimers can impact sustainable composite processing. Full article

Background: Bladder cancer (BLCA) is a malignant growth that originates from the urothelial lining of the urinary bladder. Diagnosing BLCA is complex due to the variety of tumor features and its heterogeneous nature, which leads to significant morbidity and mortality. Understanding tumor histopathology is crucial for developing tailored therapies and improving patient outcomes. Objectives: Early diagnosis and treatment are essential to lower the mortality rate associated with bladder cancer. Manual classification of muscular tissues by pathologists is labor-intensive and relies heavily on experience, which can result in interobserver variability due to the similarities in cancerous cell morphology. Traditional methods for analyzing endoscopic images are often time-consuming and resource-intensive, making it difficult to efficiently identify tissue types. Therefore, there is a strong demand for a fully automated and reliable system for classifying smooth muscle images. Methods: This paper proposes a deep learning (DL) technique utilizing the EfficientNet-B3 model and a five-fold cross-validation method to assist in the early detection of BLCA. This model enables timely intervention and improved patient outcomes while streamlining the diagnostic process, ultimately reducing both time and costs for patients. We conducted experiments using the Endoscopic Bladder Tissue Classification (EBTC) dataset for multiclass classification tasks. The dataset was preprocessed using resizing and normalization methods to ensure consistent input. In-depth experiments were carried out utilizing the EBTC dataset, along with ablation studies to evaluate the best hyperparameters. A thorough statistical analysis and comparisons with five leading DL models—ConvNeXtBase, DenseNet-169, MobileNet, ResNet-101, and VGG-16—showed that the proposed model outperformed the others. Conclusions: The EfficientNet-B3 model achieved impressive results: accuracy of 99.03%, specificity of 99.30%, precision of 97.95%, recall of 96.85%, and an F1-score of 97.36%. These findings indicate that the EfficientNet-B3 model demonstrates significant potential in accurately and efficiently diagnosing BLCA. Its high performance and ability to reduce diagnostic time and cost make it a valuable tool for clinicians in the field of oncology and urology. Full article

The increasing use of AI systems for face, object, action, scene, and emotion recognition raises significant privacy risks, particularly when processing Personally Identifiable Information (PII). Current privacy-preserving methods lack adaptability to users’ preferences and contextual requirements, and obfuscate user faces uniformly. This research proposes a user-centric, context-aware, and ontology-driven privacy protection framework that dynamically adjusts privacy decisions based on user-defined preferences, entity sensitivity, and contextual information. The framework integrates state-of-the-art recognition models for recognising faces, objects, scenes, actions, and emotions in real time on data acquired from vision sensors (e.g., cameras). Privacy decisions are directed by a contextual ontology based in Contextual Integrity theory, which classifies entities into private, semi-private, or public categories. Adaptive privacy levels are enforced through obfuscation techniques and a multi-level privacy model that supports user-defined red lines (e.g., “always hide logos”). The framework also proposes a Re-Identifiability Index (RII) using soft biometric features such as gait, hairstyle, clothing, skin tone, age, and gender, to mitigate identity leakage and to support fallback protection when face recognition fails. The experimental evaluation relied on sensor-captured datasets, which replicate real-world image sensors such as surveillance cameras. User studies confirmed that the framework was effective, with over 85.2% of participants rating the obfuscation operations as highly effective, and the other 14.8% stating that obfuscation was adequately effective. Amongst these, 71.4% considered the balance between privacy protection and usability very satisfactory and 28% found it satisfactory. GPU acceleration was deployed to enable real-time performance of these models by reducing frame processing time from 1200 ms (CPU) to 198 ms. This ontology-driven framework employs user-defined red lines, contextual reasoning, and dual metrics (RII/IVI) to dynamically balance privacy protection with scene intelligibility. Unlike current anonymisation methods, the framework provides a real-time, user-centric, and GDPR-compliant method that operationalises privacy-by-design while preserving scene intelligibility. These features make the framework appropriate to a variety of real-world applications including healthcare, surveillance, and social media. Full article

There are various mechanisms through which water droplets can be present in compressor flows, e.g., rain ingestion in aeroengines or overspray fogging used in heavy-duty gas turbines to boost power output. For the latter, droplet evaporation within the compressor leads to a cooling of the flow as well as to a shift in the fluid properties, which is beneficial to the overall process. However, due to their inertia, the majority of droplets are deposited in the first stages of a multistage compressor. While this phenomenon is generally considered in CFD computations of droplet-laden flows, the subsequent re-entrainment of collected water, the formation of new droplets, and the impact on the overall evaporation are mostly neglected because of the additional computational effort required, especially with regard to the modeling of films formed by the deposited water. The work presented here shows an approach that allows for the integration of the process of droplet deposition and re-entrainment based on relatively simple correlations and experimental observations from the literature. Thus, the two-phase flow in multistage compressors can be modelled and analyzed very efficiently. In this paper, the models and assumptions used are described first, then the results of a study performed based on a generic multistage compressor are presented, whereby the various models are integrated step by step to allow an assessment of their impact on the droplet evaporation throughout the compressor and overall performance. It can be shown that evaporation becomes largely independent of droplet size when deposition on both rotor and stator and subsequent re-entrainment of collected water is considered. In addition, open issues with regard to the future improvement of models and correlations of two-phase flow phenomena are highlighted based on the results of the current investigation. Full article

To address the limitations of traditional cotton leaf nitrogen content estimation methods, which include low efficiency, high cost, poor portability, and challenges in vegetation index acquisition owing to environmental interference, this study focused on emerging non-destructive nutrient estimation technologies. This study proposed an innovative method that integrates multi-color space fusion with deep and machine learning to estimate cotton leaf nitrogen content using smartphone-captured digital images. A dataset comprising smartphone-acquired cotton leaf images was processed through threshold segmentation and preprocessing, then converted into RGB, HSV, and Lab color spaces. The models were developed using deep-learning architectures including AlexNet, VGGNet-11, and ResNet-50. The conclusions of this study are as follows: (1) The optimal single-color-space nitrogen estimation model achieved a validation set R² of 0.776. (2) Feature-level fusion by concatenation of multidimensional feature vectors extracted from three color spaces using the optimal model, combined with an attention learning mechanism, improved the validation R² to 0.827. (3) Decision-level fusion by concatenating nitrogen estimation values from optimal models of different color spaces into a multi-source decision dataset, followed by machine learning regression modeling, increased the final validation R² to 0.830. The dual fusion method effectively enabled rapid and accurate nitrogen estimation in cotton crops using smartphone images, achieving an accuracy 5–7% higher than that of single-color-space models. The proposed method provides scientific support for efficient cotton production and promotes sustainable development in the cotton industry. Full article

Automated quality control is critical in modern manufacturing, especially for metallic cast components, where fast and accurate surface defect detection is required. This study evaluates classical Machine Learning (ML) algorithms using extracted statistical parameters and deep learning (DL) architectures including ResNet50, Capsule Networks, and a 3D Convolutional Neural Network (CNN3D) using 3D image inputs. Using the Dataset Original, ML models with the selected parameters achieved high performance: RF reached 99.4 ± 0.2% precision and 99.4 ± 0.2% sensitivity, GB 96.0 ± 0.2% precision and 96.0 ± 0.2% sensitivity. ResNet50 trained with extracted parameters reached 98.0 ± 1.5% accuracy and 98.2 ± 1.7% F1-score. Capsule-based architectures achieved the best results, with ConvCapsuleLayer reaching 98.7 ± 0.2% accuracy and 100.0 ± 0.0% precision for the normal class, and 98.9 ± 0.2% F1-score for the affected class. CNN3D applied on 3D image inputs reached 88.61 ± 1.01% accuracy and 90.14 ± 0.95% F1-score. Using the Dataset Expanded with ML and PCA-selected features, Random Forest achieved 99.4 ± 0.2% precision and 99.4 ± 0.2% sensitivity, K-Nearest Neighbors 99.2 ± 0.0% precision and 99.2 ± 0.0% sensitivity, and SVM 99.2 ± 0.0% precision and 99.2 ± 0.0% sensitivity, demonstrating consistent high performance. All models were evaluated using repeated train-test splits to calculate averages of standard metrics (accuracy, precision, recall, F1-score), and processing times were measured, showing very low per-image execution times (as low as $3.69\times {10}^{-4}$ s/image), supporting potential real-time industrial application. These results indicate that combining statistical descriptors with ML and DL architectures provides a robust and scalable solution for automated, non-destructive surface defect detection, with high accuracy and reliability across both the original and expanded datasets. Full article