Search Results (3,623)

The present study explores the application of RNNs for the prediction and propagation of flood waves along a section of the Bârsa River, Romania, as a fast alternative to classical hydraulic models, aiming to identify new ways to alert the population. Five neural architectures were analyzed as follows: S-RNN, LSTM, GRU, Bi-LSTM, and Bi-GRU. The input data for the neural networks were derived from 2D hydraulic simulations conducted using HEC-RAS software, which provided the necessary training data for the models. It should be mentioned that the input data for the hydraulic model are synthetic hydrographs, derived from the statistical processing of recorded floods. Performance evaluation was based on standard metrics such as NSE, $R^2$ MSE, and RMSE. The results indicate that all studied networks performed well, with NSE and $R^2$ values close to 1, thus validating their capacity to reproduce complex hydrological dynamics. Overall, all models yielded satisfactory results, making them useful tools particularly the GRU and Bi-GRU architectures, which showed the most balanced behavior, delivering low errors and high stability in predicting peak discharge, water level, and flood wave volume. The GRU and Bi-GRU networks yielded the best performance, with RMSE values below 1.45, MAE under 0.3, and volume errors typically under 3%. On the other hand, LSTM architecture exhibited the most significant instability and errors, especially in estimating the flood wave volume, often having errors exceeding 9% in some sections. The study concludes by identifying several limitations, including the heavy reliance on synthetic data and its local applicability, while also proposing solutions for future analyses, such as the integration of real-world data and the expansion of the methodology to diverse river basins thus providing greater significance to RNN models. The final conclusions highlight that RNNs are powerful tools in flood risk management, contributing to the development of fast and efficient early warning systems for extreme hydrological and meteorological events. Full article

In this study, we introduce an enhanced hybrid Autoencoder–Dense–Transformer Neural Network (AE-DTNN) model for developing an effective intrusion detection system (IDS) aimed at improving the performance and robustness of threat detection strategies within a rapidly changing and increasingly complex network landscape. The Autoencoder component restructures network traffic data, while a stack of Dense layers performs feature extraction to generate more meaningful representations. The Transformer network then facilitates highly precise and comprehensive classification. Our strategy incorporates adaptive synthetic sampling (ADASYN) for both binary and multi-class classification tasks, complemented by the edited nearest neighbors (ENN) technique and the use of class weights to mitigate class imbalance issues. In experiments conducted on the NF-BoT-IoT-v2 dataset, the AE-DTNN-based IDS achieved outstanding performance, with 99.98% accuracy in binary classification and 98.30% in multi-class classification. On the NSL-KDD dataset, the model reached 98.57% accuracy for binary classification and 97.50% for multi-class classification. Additionally, the model attained 99.92% and 99.78% accuracy in binary and multi-class classification, respectively, on the CSE-CIC-IDS2018 dataset. These results demonstrate the exceptional effectiveness of the proposed model in contrast to conventional approaches, highlighting its strong potential to detect a broad range of network intrusions with high reliability. Full article

The rapid proliferation of Internet of Things (IoT) devices has significantly increased vulnerability to Distributed Denial of Service (DDoS) attacks, which can severely disrupt network operations. DDoS attacks in IoT networks disrupt communication and compromise service availability, causing severe operational and economic losses. In this paper, we present a Deep Learning (DL)-based Intrusion Detection System (IDS) tailored for IoT environments. Our system employs three architectures—Convolutional Neural Networks (CNNs), Deep Neural Networks (DNNs), and Transformer-based models—to perform binary, three-class, and 12-class classification tasks on the CiC IoT 2023 dataset. Data preprocessing includes log normalization to stabilize feature distributions and SMOTE-based oversampling to mitigate class imbalance. Experiments on the CIC-IoT 2023 dataset show that, in the binary classification task, the DNN achieved 99.2% accuracy, the CNN 99.0%, and the Transformer 98.8%. In three-class classification (benign, DDoS, and non-DDoS), all models attained near-perfect performance (approximately 99.9–100%). In the 12-class scenario (benign plus 12 attack types), the DNN, CNN, and Transformer reached 93.0%, 92.7%, and 92.5% accuracy, respectively. The high precision, recall, and ROC-AUC values corroborate the efficacy and generalizability of our approach for IoT DDoS detection. Comparative analysis indicates that our proposed IDS outperforms state-of-the-art methods in terms of detection accuracy and efficiency. These results underscore the potential of integrating advanced DL models into IDS frameworks, thereby providing a scalable and effective solution to secure IoT networks against evolving DDoS threats. Future work will explore further enhancements, including the use of deeper Transformer architectures and cross-dataset validation, to ensure robustness in real-world deployments. Full article

Teledermatology relies on digital transfer of dermatological images, but compression and resolution differences compromise diagnostic quality. Image enhancement techniques are crucial to compensate for these differences and improve quality for both clinical assessment and AI-based analysis. We developed a customized image degradation pipeline simulating common artifacts in dermatological images, including blur, noise, downsampling, and compression. This synthetic degradation approach enabled effective training of DermaSR-GAN, a super-resolution generative adversarial network tailored for dermoscopic images. The model was trained on 30,000 high-quality ISIC images and evaluated on three independent datasets (ISIC Test, Novara Dermoscopic, PH²) using structural similarity and no-reference quality metrics. DermaSR-GAN achieved statistically significant improvements in quality scores across all datasets, with up to 23% enhancement in perceptual quality metrics (MANIQA). The model preserved diagnostic details while doubling resolution and surpassed existing approaches, including traditional interpolation methods and state-of-the-art deep learning techniques. Integration with downstream classification systems demonstrated up to 14.6% improvement in class-specific accuracy for keratosis-like lesions compared to original images. Synthetic degradation represents a promising approach for training effective super-resolution models in medical imaging, with significant potential for enhancing teledermatology applications and computer-aided diagnosis systems. Full article

Polarimetric Synthetic Aperture Radar (PolSAR) image classification is one of the most important applications in remote sensing. The impressive superpixel generation approaches can improve the efficiency of the subsequent classification task and restrain the influence of the speckle noise to an extent. Most of the classical PolSAR superpixel generation approaches use the features extracted manually and even only consider the pseudocolor images. They do not make full use of polarimetric information and do not necessarily lead to good enough superpixels. The deep learning methods can extract effective deep features but they are difficult to combine with superpixel generation to achieve true end-to-end training. Addressing the above issues, this study proposes an end-to-end fully convolutional superpixel generation network for PolSAR images. It integrates the extraction of polarization information features and the generation of PolSAR superpixels into one step. PolSAR superpixels can be generated based on deep polarization feature extraction and need no traditional clustering process. Both the performance and efficiency of generations of PolSAR superpixels can be enhanced effectively. The experimental results on various PolSAR datasets show that the proposed method can achieve impressive superpixel segmentation by fitting the real boundaries of different types of ground objects effectively and efficiently. It can achieve excellent classification performance by connecting a very simple classification network, which is helpful to improve the efficiency of the subsequent PolSAR image classification tasks. Full article

Ensuring bacterial safety of blood transfusions remains a critical focus in medicine. We investigated a novel pathogen reduction technology utilizing nylon functionalized with synthetic dyes (nylon affinity networks) to capture and remove bacteria from plasma. In the initial screening process, we spiked phosphate buffer solution (PBS) and human plasma (1 mL each) with 10 or 100 colony forming units (cfu) of either Escherichia coli or Staphylococcus epidermidis, exposed the suspensions to affinity networks and assessed the extent of bacterial reduction using agar plate cultures as the assay output. Nineteen synthetic dyes were tested. Among these, Alcian Blue exhibited the best performance with both bacterial strains in both PBS and plasma. Next, bacterial suspensions of approximately 1 and 2 cfu/mL in 10 and 50 mL, respectively, were treated with Alcian Blue affinity networks in three sequential capture steps. This procedure resulted in complete bacterial depletion, as demonstrated by the lack of bacterial growth in the remaining fraction. The viability of the captured bacteria was confirmed by plating the post-treatment affinity networks on agar. Alcian Blue affinity networks captured and sequestered a few plasma proteins identified by liquid chromatography tandem mass spectrometry. These findings support the potential applicability of nylon affinity networks to enhance transfusion safety, although additional investigations are needed. Full article

The number of CNN application areas is growing, which leads to the need for training data. The research conducted in this work aimed to obtain effective detection models trained only using simplified synthetic objects (SSOs). The research was conducted on inland shallow water areas, while images of bottom objects were obtained using a UAV platform. The work consisted in preparing SSOs, thanks to which composite images were created. On such training data, 120 models based on the YOLO (You Only Look Once) network were obtained. The study confirmed the effectiveness of models created using YOLOv3, YOLOv5, YOLOv8, YOLOv9, and YOLOv10. A comparison was made between versions of YOLO. The influence of the amount of training data, SSO type, and augmentation parameters used in the training process was analyzed. The main parameter of model performance was the F1-score. The calculated statistics of individual models indicate that the most effective networks use partial augmentation, trained on sets consisting of 2000 SSOs. On the other hand, the increased transparency of SSOs resulted in increasing the diversity of training data and improving the performance of models. This research is developmental, and further research should improve the processes of obtaining detection models using deep networks. Full article

Privacy preservation poses significant challenges in third-party data sharing, particularly when handling table data containing personal information such as demographic and behavioral records. Synthetic table data generation has emerged as a promising solution to enable data analysis while mitigating privacy risks. While Generative Adversarial Networks (GANs) are widely used for this purpose, they exhibit limitations in modeling table data due to challenges in handling mixed data types (numerical/categorical), non-Gaussian distributions, and imbalanced variables. To address these limitations, this study proposes a novel adversarial learning framework integrating gradient boosting trees for synthesizing table data, called Adversarial Gradient Boosting Decision Tree (AGBDT). Experimental evaluations on several datasets demonstrate that our method outperforms representative baseline models regarding statistical similarity and machine learning utility. Furthermore, we introduce a privacy-aware adaptation of the framework by incorporating k-anonymization constraints, effectively reducing overfitting to source data while maintaining practical usability. The results validate the balance between data utility and privacy preservation achieved by our approach. Full article

To address the issues in existing 3D point cloud facet generation networks, specifically, the tendency to produce a large number of empty facets and the uncertainty in facet count, this paper proposes a novel deep learning framework for robust facet segmentation. Based on the generated facet set, two exploratory applications are further developed. First, to overcome the bottleneck where inaccurate empty-facet detection impairs the downsampling performance, a facet-abstracted downsampling method is introduced. By using a learned facet classifier to filter out and discard empty facets, retaining only non-empty surface facets, and fusing point coordinates and local features within each facet, the method achieves significant compression of point cloud data while preserving essential geometric information. Second, to solve the insufficient precision in organ segmentation within crop point clouds, a facet growth-based segmentation algorithm is designed. The network first predicts the edge scores for the facets to determine the seed facets. The facets are then iteratively expanded according to adjacent-facet similarity until a complete organ region is enclosed, thereby enhancing the accuracy of segmentation across semantic boundaries. Finally, the proposed facet segmentation network is trained and validated using a synthetic dataset. Experiments show that, compared with traditional methods, the proposed approach significantly outperforms both downsampling accuracy and instance segmentation performance. In various crop scenarios, it demonstrates excellent geometric fidelity and semantic consistency, as well as strong generalization ability and practical application potential, providing new ideas for in-depth applications of facet-level features in 3D point cloud analysis. Full article

Stoichiometric single crystals of Mn₄Al₁₁ were synthesized from the elements using Sn as a flux. The crystal structure of Mn₄Al₁₁ was investigated using single crystal X-ray diffraction and showed a complex triclinic structure with a relatively small unit cell and interpenetrating networks of Mn and Al atoms. While our results generally agree with the previously reported data in the basic structure features such as triclinic symmetry and structure type, the atomic parameters differ significantly, likely due to different synthetic techniques producing off-stoichiometry or doped crystals used in the previous works. Our structural analysis showed that the view of the Mn substructure as isolated zigzag chains is incomplete. Instead, the Mn chains are coupled in corrugated layers by long Mn-Mn bonds. The high quality of the crystals with the stoichiometric composition also enabled us to study magnetic behavior in great detail and reveal previously unobserved magnetic ordering. Our magnetization measurements showed that Mn₄Al₁₁ is an antiferromagnet with T_N of 65 K. The presence of the maximum above T_N also suggests strong local interactions indicative of low-dimensional magnetic behavior, which likely stems from lowered dimensionality of the Mn substructure. Full article

Diabetes mellitus (DM) is one of the leading causes of death and disability worldwide and its prevalence continues to rise. Chronic hyperglycemia exposes patients to severe complications. Among these, diabetic vascular lesions are the most destructive. Their primary driver is the synergistic interaction between hyperglycemia-induced oxidative stress and chronic inflammation. This review systematically elucidates how multiple pathological pathways—namely, metabolic dysregulation, mitochondrial dysfunction, endoplasmic reticulum stress, and epigenetic reprogramming—cooperate to drive oxidative stress and inflammatory cascades. Confronting this complex pathological network, natural products, unlike conventional single-target synthetic drugs, exert multi-target synergistic effects, simultaneously modulating several key pathogenic networks. This enables the restoration of redox homeostasis and the suppression of inflammatory responses, thereby improving vascular function and delaying both microvascular and macrovascular disease progression. However, the clinical translation of natural products still faces multiple challenges and requires comprehensive mechanistic studies and rigorous validation to fully realize their therapeutic potential. Full article

Hydrogels, characterized by their high water content, tunable mechanical properties, and excellent biocompatibility, have emerged as a promising material platform for the preservation of cultural heritage. Their unique physicochemical features enable non-invasive and adaptable solutions for environmental regulation, structural stabilization, and antifungal protection. This review provides a comprehensive overview of recent progress in hydrogel-based strategies specifically developed for the conservation of cultural relics, with a particular focus on antifungal performance—an essential factor in preventing biodeterioration. Current hydrogel systems, composed of natural or synthetic polymer networks integrated with antifungal agents, demonstrate the ability to suppress fungal growth, regulate humidity, alleviate mechanical stress, and ensure minimal damage to artifacts during application. This review also highlights future research directions, such as the application prospects of novel materials, including stimuli-responsive hydrogels and self-dissolving hydrogels. As an early exploration of the use of hydrogels in antifungal protection and broader cultural heritage conservation, this work is expected to promote the wider application of this emerging technology, contributing to the effective preservation and long-term transmission of cultural heritage worldwide. Full article

Wet areas, which are locations in the landscape that consistently retain moisture, and channel networks are important landscape compartments, with key hydrological and ecological functions. Hence, defining their spatial boundaries is an important step towards sustainable watershed management. In catchments of developing countries, wet areas and small order channels of river networks are rarely mapped, although they represent a crucial component of local livelihoods and ecosystems. In this study, topographic attributes generated with a 30 m SRTM DEM were used to map wet areas and stream networks of two tropical catchments in Central Brazil. The topographic attributes for wet areas were the local slope and the slope curvature, and the Topographic Wetness Index (TWI) was used to delineate the stream networks. Threshold values of the selected topographic attributes were calibrated in the Santa Maria catchment, comparing the synthetically generated wet areas and drainage networks with corresponding reference (map) features, and validated in the nearby Santa Maria basin. Drainage network and wet area delineation accuracies were estimated using random basin transects and multi-criteria and confusion matrix methods. The drainage network accuracies were 67.2% and 70.7%, and wet area accuracies were 72.7% and 73.8%, for the Santa Maria and Gama catchments, respectively, being equivalent or higher than previous studies. The mapping errors resulted from model incompleteness, DEM vertical inaccuracy, and cartographic misrepresentation of the reference topographic maps. The study’s novelty is the use of readily available information to map, with simplicity and robustness, wet areas and channel initiation in data-scarce, tropical environments. Full article

Microalgae, with their unparalleled capabilities for sunlight-driven growth, CO₂ fixation, and synthesis of diverse high-value compounds, represent sustainable cell factories for a circular bioeconomy. However, industrial deployment has been hindered by biological constraints and the inadequacy of conventional genetic tools. The advent of CRISPR-Cas systems initially provided precise gene editing via targeted DNA cleavage. This review argues that the true transformative potential lies in moving decisively beyond cutting to harness CRISPR as a versatile synthetic biology “Swiss Army Knife”. We synthesize the rapid evolution of CRISPR-derived tools—including transcriptional modulators (CRISPRa/i), epigenome editors, base/prime editors, multiplexed systems, and biosensor-integrated logic gates—and their revolutionary applications in microalgal engineering. These tools enable tunable gene expression, stable epigenetic reprogramming, DSB-free nucleotide-level precision editing, coordinated rewiring of complex metabolic networks, and dynamic, autonomous control in response to environmental cues. We critically evaluate their deployment to enhance photosynthesis, boost lipid/biofuel production, engineer high-value compound pathways (carotenoids, PUFAs, proteins), improve stress resilience, and optimize carbon utilization. Persistent challenges—species-specific tool optimization, delivery efficiency, genetic stability, scalability, and biosafety—are analyzed, alongside emerging solutions and future directions integrating AI, automation, and multi-omics. The strategic integration of this CRISPR toolkit unlocks the potential to engineer robust, high-productivity microalgal cell factories, finally realizing their promise as sustainable platforms for next-generation biomanufacturing. Full article

Environmental noise is inevitable during seismic data acquisition, with major sources including heavy machinery, rivers, wind, and other environmental factors. During field data acquisition, it is important to assess the impact of environmental noise and evaluate data quality. In subsequent seismic data processing, these noise components also need to be eliminated. Accurate identification of noise traces facilitates rapid quality control (QC) during fieldwork and provides a reliable basis for targeted noise attenuation. Conventional environmental noise identification primarily relies on amplitude differences. However, in seismic data, high-amplitude signals are not necessarily caused by environmental noise. For example, surface waves or traces near the shot point may also exhibit high amplitudes. Therefore, relying solely on amplitude-based criteria has certain limitations. To improve noise identification accuracy, we use the Mel-spectrogram to extract features from seismic data and construct the dataset. Compared to raw time-series signals, the Mel-spectrogram more clearly reveals energy variations and frequency differences, helping to identify noise traces more accurately. We then employ a Vision Transformer (ViT) network to train a model for identifying noise in seismic data. Tests on synthetic and field data show that the proposed method performs well in identifying noise. Moreover, a denoising case based on synthetic data further confirms its general applicability, making it a promising tool in seismic data QC and processing workflows. Full article