Search Results (704)

Sewer systems are essential for sustainable infrastructure management, influencing environmental, social, and economic aspects. However, sewer network capacity is under significant pressure, with many systems overwhelmed by challenges such as climate change, ageing infrastructure, and increasing inflow and infiltration, particularly through groundwater infiltration (GWI). Current research in this area has primarily focused on general sewer performance, with limited attention to high-resolution, spatially explicit assessments of sewer exposure to GWI, highlighting a critical knowledge gap. This study responds to this gap by developing a high-resolution GWI assessment. This is achieved by integrating fuzzy-analytical hierarchy process (AHP) with geographic information systems (GISs) and machine learning (ML) to generate GWI probability maps across the Dawlish region, southwest United Kingdom, complemented by sensitivity analysis to identify the key drivers of sewer network vulnerability. To this end, 16 hydrological–hydrogeological thematic layers were incorporated: elevation, slope, topographic wetness index, rock, alluvium, soil, land cover, made ground, fault proximity, fault length, mass movement, river proximity, flood potential, drainage order, groundwater depth (GWD), and precipitation. A GWI probability index, ranging from 0 to 1, was developed for each 1 m × 1 m area per season. The model domain was then classified into high-, intermediate-, and low-GWI-risk zones using K-means clustering. A consistency ratio of 0.02 validated the AHP approach for pairwise comparisons, while locations of storm overflow (SO) discharges and model comparisons verified the final outputs. SOs predominantly coincided with areas of high GWI probability and high-risk zones. Comparison of AHP-weighted GIS output clustered via K-means with direct K-means clustering of AHP-weighted layers yielded a Kappa value of 0.70, with an 81.44% classification match. Sensitivity analysis identified five key factors influencing GWI scores: GWD, river proximity, flood potential, rock, and alluvium. The findings underscore that proxy-based geospatial and machine learning approaches offer an effective and scalable method for mapping sewer network exposure to GWI. By enabling high-resolution risk assessment, the proposed framework contributes a novel proxy and machine-learning-based screening tool for the management of smart cities. This supports predictive maintenance, optimised infrastructure investment, and proactive management of GWI in sewer networks, thereby reducing costs, mitigating environmental impacts, and protecting public health. In this way, the method contributes not only to improved sewer system performance but also to advancing the sustainability and resilience goals of smart cities. Full article

This research considers an Artificial Intelligence (AI)-driven omnichannel logistics network for bioethanol supply from Russia to China. As a renewable, low-carbon transport fuel, bioethanol plays a critical role in energy diversification and decarbonization strategies for both Russia and China. However, its flammability and temperature sensitivity impose stringent requirements on transport infrastructure and supply chain management, making it a typical application scenario for exploring intelligent logistics models. The proposed model integrates information, transportation, and financial flows into a unified simulation framework designed to support flexible and sustainable cross-border (CB) logistics. Using a combination of machine learning, multi-objective evaluation, and reinforcement learning (RL), the system models and ranks alternative transportation routes under varying operational conditions. Results indicate that the mixed corridor through Kazakhstan and Kyrgyzstan achieves the best overall balance of cost, time, emissions, and customs reliability, outperforming single-country routes. The findings highlight the potential of AI-enhanced logistics systems in supporting low-carbon energy trade and CB infrastructure coordination. Full article

Neurogenic bladder is a debilitating condition caused by neurological dysfunction that impairs urinary control, often requiring timed intermittent catheterisation. Although effective, intermittent catheterisation is invasive, uncomfortable, and associated with infection risks, reducing patients’ quality of life. SafeBladder is a low-cost wearable device developed to enable real-time, non-invasive bladder volume monitoring using near-infrared spectroscopy (NIRS) and machine learning algorithms. The prototype employs LEDs and photodetectors to measure light attenuation through abdominal tissues. Bladder filling was simulated through experimental tests using stepwise water additions to containers and tissue-mimicking phantoms, including silicone and porcine tissue. Machine learning models, including Linear Regression, Support Vector Regression, and Random Forest, were trained to predict volume from sensor data. The results showed the device is sensitive to volume changes, though ambient light interference affected accuracy, suggesting optimal use under clothing or in low-light conditions. The Random Forest model outperformed others, with a Mean Absolute Error (MAE) of 25 ± 4 mL and $R^2$ of 0.90 in phantom tests. These findings support SafeBladder as a promising, non-invasive solution for bladder monitoring, with clinical potential pending further calibration and validation in real-world settings. Full article

Owing to its excellent crack resistance and durability, High-Performance Fiber-Reinforced Concrete (HPFRC) has been extensively applied in engineering structures exposed to extreme loading conditions. The Mode I dynamic fracture strength of HPFRC under high-strain-rate conditions exhibits significant strain-rate sensitivity and nonlinear response characteristics. However, existing experimental methods for strength measurement are limited by high costs and the absence of standardized testing protocols. Meanwhile, conventional data-driven models for strength prediction struggle to achieve both high-precision prediction and physical interpretability. To address this, this study introduces a dynamic fracture strength prediction method based on a feature-weighted linear ensemble (FWL) mechanism. A comprehensive database comprising 161 sets of high-strain-rate test data on HPFRC fracture strength was first constructed. Key modeling variables were then identified through correlation analysis and an error-driven feature selection approach. Subsequently, six representative machine learning models (KNN, RF, SVR, LGBM, XGBoost, MLPNN) were employed as base learners to construct two types of ensemble models, FWL and Voting, enabling a systematic comparison of their performance. Finally, the predictive mechanisms of the models were analyzed for interpretability at both global and local scales using SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) methods. The results demonstrate that the FWL model achieved optimal predictive performance on the test set (R² = 0.908, RMSE = 2.632), significantly outperforming both individual models and the conventional ensemble method. Interpretability analysis revealed that strain rate and fiber volume fraction are the primary factors influencing dynamic fracture strength, with strain rate demonstrating a highly nonlinear response mechanism across different ranges. The integrated prediction framework developed in this study offers the combined advantages of high accuracy, robustness, and interpretability, providing a novel and effective approach for predicting the fracture behavior of HPFRC under high-strain-rate conditions. Full article

Background/Objectives: Bones are essential components of the human body, providing structural support, enabling mobility, storing minerals, and protecting internal organs. Bone fractures (BFs) are common injuries that result from excessive physical force and can lead to serious complications, including bleeding, infection, impaired oxygenation, and long-term disability. Early and accurate identification of fractures through radiographic imaging is critical for effective treatment and improved patient outcomes. However, manual evaluation of X-rays is often time-consuming and prone to diagnostic errors due to human limitations. To address this, artificial intelligence (AI), particularly deep learning (DL), has emerged as a powerful tool for enhancing diagnostic precision in medical imaging. Methods: This research introduces a novel convolutional neural network (CNN) model, the Multi-Level Feature Fusion Network (MLFNet), designed to capture and integrate both low-level and high-level image features. The model was evaluated using the Bone Fracture Multi-Region X-ray (BFMRX) dataset. Preprocessing steps included image normalization, resizing, and contrast enhancement to ensure stable convergence, reduce sensitivity to lighting variations in radiographic images, and maintain consistency. Ablation studies were conducted to assess architectural variations, confirming the model’s robustness and generalizability across data distributions. MLFNet’s high accuracy, interpretability, and efficiency make it a promising solution for clinical deployment. Results: MLFNet achieved an impressive accuracy of 99.60% as a standalone model and 98.81% when integrated into hybrid ensemble architectures with five leading pre-trained DL models. Conclusions: The proposed approach supports timely and precise fracture detection, optimizing the diagnostic process and reducing healthcare costs. This approach offers significant potential to aid clinicians in fields such as orthopedics and radiology, contributing to more equitable and effective patient care. Full article

Autonomous driving represents a transformative advancement with the potential to significantly impact daily mobility, including enabling independent vehicle operation for individuals with visual disabilities. The commercialization of autonomous driving requires guaranteed safety and accuracy, underscoring the need for robust localization and environmental perception algorithms. In cost-sensitive platforms such as delivery robots and electric vehicles, cameras are increasingly favored for their ability to provide rich visual information at low cost. Despite recent progress, existing visual–inertial odometry systems still suffer from degraded accuracy in challenging conditions, which limits their reliability in real-world autonomous navigation scenarios. Estimating 3D positional changes using only 2D image sequences remains a fundamental challenge primarily due to inherent scale ambiguity and the presence of dynamic scene elements. In this paper, we present a visual–inertial odometry framework incorporating a depth estimation model trained without ground-truth depth supervision. Our approach leverages a self-supervised learning pipeline enhanced with knowledge distillation via foundation models, including both self-distillation and geometry-aware distillation. The proposed method improves depth estimation performance and consequently enhances odometry estimation without modifying the network architecture or increasing the number of parameters. The effectiveness of the proposed method is demonstrated through comparative evaluations on both the public KITTI dataset and a custom campus driving dataset, showing performance improvements over existing approaches. Full article

In Directed Energy Deposition (DED), modeling the molten pool temperature field is crucial for precise temperature control, process optimization, and quality improvement. However, conventional numerical methods suffer from limitations such as high computational costs and poor transferability. This study proposes a physics-informed neural network with dynamic learning rate (DLR-PINN) model, which integrates transfer learning to enable rapid prediction of 3D temperature fields and dimensions of molten pools across process parameters. Its validity is verified by a finite element method (FEM) calibrated via single-track DED experiments. Results show that DLR-PINN exhibits superior convergence and stability compared to traditional PINN. Combined with transfer learning, training efficiency is significantly enhanced, with a single prediction taking only 10 s. Using the FEM as the benchmark, it achieves a mean absolute percentage error (MAPE) of 0.53% for temperature prediction, and MAPE of 3.69%, 2.48%, and 6.96% for molten pool dimension predictions, respectively. Sensitivity analysis of process parameters reveals that scanning speed has a significantly greater regulatory effect on molten pool characteristics than laser power. Additionally, the temperature field of the flat-top heat source is more uniform than that of the Gaussian heat source, which is more conducive to improving printing quality and efficiency. Full article

Optical coherence tomography (OCT) is a leading imaging technique for diagnosing retinal disorders such as age-related macular degeneration and diabetic retinopathy. Its ability to detect structural changes, especially in the optic nerve head, has made it vital for early diagnosis and monitoring. This paper surveys techniques for ocular disease prediction using OCT, focusing on both hand-crafted and deep learning-based feature extractors. While the field has seen rapid growth, a detailed comparative analysis of these methods has been lacking. We address this by reviewing research from the past 20 years, evaluating methods based on accuracy, sensitivity, specificity, and computational cost. Key diseases examined include glaucoma, diabetic retinopathy, cataracts, amblyopia, and macular degeneration. We also assess public OCT datasets widely used in model development. A unique contribution of this paper is the exploration of adversarial attacks targeting OCT-based diagnostic systems and the vulnerabilities of different feature extraction techniques. We propose a practical, robust defense strategy that integrates with existing models and outperforms current solutions. Our findings emphasize the value of combining classical and deep learning methods with strong defenses to enhance the security and reliability of OCT-based diagnostics, and we offer guidance for future research and clinical integration. Full article

Diabetic retinopathy (DR), a critical ocular disease that can lead to blindness, demands early and accurate diagnosis to prevent vision loss. Current automated DR diagnosis methods face two core challenges: first, subtle early lesions such as microaneurysms are often missed due to insufficient feature extraction; second, there is a persistent trade-off between model accuracy and efficiency—lightweight architectures often sacrifice precision for real-time performance, while high-accuracy models are computationally expensive and difficult to deploy on resource-constrained edge devices. To address these issues, this study presents a novel deep learning framework integrating depthwise separable convolution and a multi-view attention mechanism (MVAM) for efficient DR diagnosis using retinal images. The framework employs multi-scale feature fusion via parallel 3 × 3 and 5 × 5 convolutions to capture lesions of varying sizes and incorporates Gabor filters to enhance vascular texture and directional lesion modeling, improving sensitivity to early structural abnormalities while reducing computational costs. Experimental results on both the diabetic retinopathy (DR) dataset and ocular disease (OD) dataset demonstrate the superiority of the proposed method: it achieves a high accuracy of 0.9697 on the DR dataset and 0.9669 on the OD dataset, outperforming traditional methods such as CNN_eye, VGG, and UNet by more than 1 percentage point. Moreover, its training time is only half that of U-Net (on DR dataset) and VGG (on OD dataset), highlighting its potential for clinical DR screening. Full article

In maritime blurred image recognition, ship collision accidents frequently result from three primary blur types: (1) motion blur from vessel movement in complex sea conditions, (2) defocus blur due to water vapor refraction, and (3) scattering blur caused by sea fog interference. This paper proposes a dual-branch recognition method specifically designed for motion blur, which represents the most prevalent blur type in maritime scenarios. Conventional approaches exhibit constrained computational efficiency and limited adaptability across different modalities. To overcome these limitations, we propose a hybrid CNN–Transformer architecture: the CNN branch captures local blur characteristics, while the enhanced Transformer module models long-range dependencies via attention mechanisms. The CNN branch employs a lightweight ResNet variant, in which conventional residual blocks are substituted with Multi-Scale Gradient-Aware Residual Block (MSG-ARB). This architecture employs learnable gradient convolution for explicit local gradient feature extraction and utilizes gradient content gating to strengthen blur-sensitive region representation, significantly improving computational efficiency compared to conventional CNNs. The Transformer branch incorporates a Hierarchical Swin Transformer (HST) framework with Shifted Window-based Multi-head Self-Attention for global context modeling. The proposed method incorporates blur invariant Positional Encoding (PE) to enhance blur spectrum modeling capability, while employing DyT (Dynamic Tanh) module with learnable α parameters to replace traditional normalization layers. This architecture achieves a significant reduction in computational costs while preserving feature representation quality. Moreover, it efficiently computes long-range image dependencies using a compact 16 × 16 window configuration. The proposed feature fusion module synergistically integrates CNN-based local feature extraction with Transformer-enabled global representation learning, achieving comprehensive feature modeling across different scales. To evaluate the model’s performance and generalization ability, we conducted comprehensive experiments on four benchmark datasets: VAIS, GoPro, Mini-ImageNet, and Open Images V4. Experimental results show that our method achieves superior classification accuracy compared to state-of-the-art approaches, while simultaneously enhancing inference speed and reducing GPU memory consumption. Ablation studies confirm that the DyT module effectively suppresses outliers and improves computational efficiency, particularly when processing low-quality input data. Full article

Stereotactic radiosurgery (SRS) for multiple brain metastases can be delivered with a single isocenter and non-coplanar arcs, achieving highly conformal dose distributions at the cost of extreme modulation of treatment machine parameters. As a result, SRS plans are at a higher risk of patient-specific quality assurance (PSQA) failure compared to standard treatments. This study aimed to develop a machine-learning (ML) model to predict the PSQA outcome (gamma passing rate, GPR) of SRS plans. Five hundred and ninety-two consecutive patients treated between 2020 and 2024 were selected. GPR analyses were performed using a 3%/1 mm criterion and a 95% action limit for each arc. Fifteen plan complexity metrics were used as input features to predict the GPR of an arc. A stratified and a time-series approach were employed to split the data into training (1555 arcs), validation (389 arcs), and test (486 arcs) sets. The ML model achieved a mean absolute error of 2.6% on the test set, with a 0.83% median residual value (measured/predicted). Lower values of the measured GPR tended to be overestimated. Sensitivity and specificity were 93% and 56%, respectively. ML models for virtual QA of SRS can be integrated into clinical practice, facilitating more efficient PSQA approaches. Full article

Accurate workload prediction is essential for proactive resource allocation in large-scale Content Delivery Networks (CDNs), where traffic patterns are highly dynamic and geographically distributed. This paper introduces a CDN-tailored prediction and autoscaling framework that integrates statistical and deep learning models within an adaptive feedback loop. The framework is evaluated using 18 months of real traffic traces from a production multi-tier CDN, capturing realistic workload seasonality, cache–tier interactions, and propagation delays. Unlike generic cloud-edge predictors, our design incorporates CDN-specific features and model-switching mechanisms to balance prediction accuracy with computational cost. Seasonal ARIMA (S-ARIMA), Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Online Sequential Extreme Learning Machine (OS-ELM) are combined to support both short-horizon scaling and longer-term capacity planning. The predictions drive a queue-based resource-estimation model, enabling proactive cache–server scaling with low rejection rates. Experimental results demonstrate that the framework maintains high accuracy while reducing computational overhead through adaptive model selection. The proposed approach offers a practical, production-tested solution for predictive autoscaling in CDNs and can be extended to other latency-sensitive edge-cloud services with hierarchical architectures. Full article

In recent years, the evolution of cyber-attacks has exposed critical vulnerabilities in conventional defense mechanisms, particularly across national infrastructure systems such as power, transportation, and finance. Attackers are increasingly deploying persistent and sophisticated techniques to exfiltrate or manipulate sensitive data, surpassing static defense methods that depend on known vulnerabilities. This growing threat landscape underscores the urgent need for more advanced and adaptive defensive strategies to counter continuously evolving attack vectors. To address this challenge, this paper proposes a novel reinforcement learning-based optimization framework integrated with a Dynamic Heterogeneous Redundancy (DHR) architecture. Our approach uniquely utilizes reinforcement learning for the dynamic scheduling of encryption-layer configurations within the DHR framework, enabling adaptive adjustment of defense policies based on system status and threat progression. We evaluate the proposed system in a simulated adversarial environment, where reinforcement learning continuously adjusts encryption strategies and defense behaviors in response to evolving attack patterns and operational dynamics. Experimental results demonstrate that our method achieves a higher defense success rate while maintaining lower defense costs, thereby enhancing system resilience against cyber threats and improving the efficiency of defensive resource allocation. Full article

Driven by growing public security demands and the advancement of intelligent surveillance systems, person re-identification (ReID) has emerged as a prominent research focus in the field of computer vision. However, this task presents challenges due to its high sensitivity to variations in visual appearance caused by factors such as body pose and camera parameters. Although deep learning-based methods have achieved marked progress in ReID, the high cost of annotation remains a challenge that cannot be overlooked. To address this, we propose an unsupervised attribute learning framework that eliminates the need for costly manual annotations while maintaining high accuracy. The framework learns the mid-level human attributes (such as clothing type and gender) that are robust to substantial visual appearance variations and can hence boost the accuracy of attributes with a small amount of labeled data. To carry out our framework, we present a part-based convolutional neural network (CNN) architecture, which consists of two components for image and body attribute learning on a global level and upper- and lower-body image and attribute learning at a local level. The proposed architecture is trained to learn attribute-semantic and identity-discriminative feature representations simultaneously. For model learning, we first train our part-based network using a supervised approach on a labeled attribute dataset. Then, we apply an unsupervised clustering method to assign pseudo-labels to unlabeled images in a target dataset using our trained network. To improve feature compatibility, we introduce an attribute consistency scheme for unsupervised domain adaptation on this unlabeled target data. During training on the target dataset, we alternately perform three steps: extracting features with the updated model, assigning pseudo-labels to unlabeled images, and fine-tuning the model. Through a unified framework that fuses complementary attribute-label and identity label information, our approach achieves considerable improvements of 10.6% and 3.91% mAP on Market-1501→DukeMTMC-ReID and DukeMTMC-ReID→Market-1501 unsupervised domain adaptation tasks, respectively. Full article

High-fidelity finite element analysis (FEA) plays a key role in structural engineering by enabling accurate simulation of displacement, stress, and internal forces under static loads. However, its high computational cost limits applicability in real-time control, iterative design, and large-scale uncertainty quantification. Surrogate modeling provides a computationally efficient alternative by learning input–output mappings from precomputed simulations. Yet, the performance of individual surrogates is often sensitive to data distribution and model assumptions. To enhance both accuracy and robustness, we propose a model averaging framework based on Jackknife Model Averaging (JMA) that integrates six surrogate models: polynomial response surfaces (PRSs), support vector regression (SVR), radial basis function (RBF) interpolation, eXtreme Gradient Boosting (XGB), Light Gradient Boosting Machine (LGBM), and Random Forest (RF). Three ensembles are formed: JMA1 (classical models), JMA2 (tree-based models), and JMA3 (all models). JMA assigns optimal convex weights using cross-validated out-of-fold errors without a meta-learner. We evaluate the framework on the Static Analysis Dataset with over 300,000 FEA simulations. Results show that JMA consistently outperforms individual models in root mean squared error, mean absolute error, and the coefficient of determination, while also producing tighter, better-calibrated conformal prediction intervals. These findings support JMA as an effective tool for surrogate-based structural analysis. Full article