Search Results (3,003)

Reliable and efficient methods to identify Parkinson’s disease (PD) from EEG signals are highly valuable for studying early disease markers and supporting timely intervention research. In this work, we propose a lightweight deep learning framework for PD detection using electroencephalogram (EEG) signals. The framework employs a novel spatial–frequency multiscale feature fusion strategy to characterize both local neural activity and global inter-channel connectivity. To achieve high accuracy with low computational cost, we design a hybrid network combining the DCFConv lightweight convolution module and the OSGAttn orthogonal global attention module. Experiments on the public Iowa PD-EEG dataset show that our method achieves 99.81% accuracy, 100% precision, and 99.62% sensitivity, while maintaining only 0.16 M parameters and 44.08 M FLOPs. The results demonstrate that the proposed approach provides an effective, efficient, and reliable solution for PD EEG signal classification. Full article

Accurate structural damage identification under limited data availability and measurement noise remains a persistent challenge in structural health monitoring (SHM). This study proposes TLCA-RATNet, a transfer learning-enhanced residual attention temporal network for vibration-based damage state classification under noisy and small sample conditions. RATNet integrates adaptive threshold residual denoising, residual attention, and bidirectional gated recurrent unit (BiGRU)-based temporal modeling to suppress noise, emphasize damage-sensitive features, and capture global temporal dependencies. The current implementation is formulated as a single-task classifier, in which local feature refinement and global temporal representation are jointly optimized end-to-end through a unified damage classification objective. Transfer learning further initializes the target domain model using knowledge learned from a data-rich source structure, while regularized fine-tuning reduces overfitting on limited target samples. Experiments were conducted on a six-story lumped-mass shear structure, a three-story physical frame, and the IASC-ASCE SHM benchmark structure, using 10–145 training samples per damage class and additive-noise conditions ranging from 0 to 30 dB signal-to-noise ratio. On Dataset 2 at 5 dB, TLCA-RATNet achieved an accuracy of 89.86%, exceeding LSTM and CNN-BiGRU by 9.94 and 13.10 percentage points, respectively. On Dataset 3 at 0 dB, it achieved 86.00% accuracy, outperforming CNN-BiGRU by 10.72 percentage points. In the limited sample transfer experiment on Dataset 2, transfer learning increased the accuracy from 93.05% to 100.00%, representing a gain of 6.95 percentage points over training from scratch. These results indicate that TLCA-RATNet provides a data-efficient and noise-robust approach for damage state screening and rapid model adaptation in SHM applications with scarce labeled data and noisy measurements. Full article

Coiling temperature is a pivotal process parameter for hot-rolled strip steel, which directly determines the microstructure and mechanical properties of final products. Affected by the coupling of multiple process variables, coiling temperature presents strong nonlinearity and complex time-varying characteristics. Traditional heat transfer mechanism models, Random Forest (RF), Extreme Learning Machine (ELM) and single Long Short-Term Memory (LSTM) networks fail to fully explore the deep correlation among variables. In addition, their hyperparameters are generally selected by manual trial-and-error, leading to unsatisfactory prediction accuracy and poor robustness in practical production. To address the above limitations, this paper proposes a novel prediction model named WOA-CNN-GRU-SE, where the Whale Optimization Algorithm (WOA) is adopted for parameter optimization. Firstly, Convolutional Neural Network (CNN) is utilized to extract local coupling features from various working condition parameters. Secondly, the Squeeze-and-Excitation (SE) attention mechanism is applied to adaptively recalibrate channel weights, which enhances key features closely related to temperature variation and suppresses redundant interference information. Afterwards, Gated Recurrent Unit (GRU) is employed to conduct in-depth learning of temporal features. Furthermore, WOA is used to globally optimize critical hyperparameters, including learning rate, the number of GRU hidden units and L2 regularization coefficient, so as to eliminate the drawbacks of manual parameter tuning. Comparative experiments are conducted on actual production data from a hot rolling line. The results demonstrate that the proposed model outperforms CNN-GRU, CNN-GRU-SE, LSTM, RF and ELM in prediction performance. Its hit rate reaches 92.56% within the industrial error range of ±6 °C. This model effectively realizes accurate prediction of coiling temperature under complex working conditions and possesses great application potential in industrial practice. Full article

Pharmaceuticals and personal care products (PPCPs) are contaminants of emerging global concern, yet their long-term fate and distribution in strategic trans-boundary water source areas remain underexplored. This study conducted a comprehensive monitoring campaign spanning from 2018 to 2020, with six sampling campaigns carried out in August 2018, December 2018, May 2019, August 2019, December 2019, and May 2020, to investigate the occurrence, spatiotemporal dynamics, source apportionment, and ecological risks of 54 PPCPs in the surface water and sediments of the Taipu River in eastern China. Utilizing non-target screening via liquid chromatography high-resolution mass spectrometry (LC-HRMS), 54 PPCPs across 15 categories were detected, with average total concentrations of 28.60 ng/L in surface water and 13.01 ng/g in sediments. Sulfonamides, quinolones, and non-steroidal anti-inflammatory drugs (NSAIDs) dominated the aqueous phase (with bisphenol A and sulfamethoxazole highly prevalent), while hormones (e.g., estriol) and quinolones (e.g., ciprofloxacin (CIP)) exhibited significant accumulation in the benthic zone. Spatiotemporal analysis revealed a seasonal pattern where PPCP concentrations were higher in the dry season than in the wet season, primarily driven by hydrological dilution and climate-induced degradation. Furthermore, Positive Matrix Factorization (EPA PMF 5.0) extracted five distinct source factors for each matrix. Specifically, surface water pollution was primarily driven by domestic wastewater, municipal effluents, medical discharges, and localized pharmaceutical emissions, whereas sediments acted as a long-term sink predominantly for domestic wastewater, agricultural/veterinary runoff, and municipal/industrial emissions. Ecological risk assessment indicated that while most of the nine high-frequency contaminants posed low risks, specific compounds such as CIP, SM2, and E3 presented medium risks in surface water, whereas ENR, CIP, and SPI posed medium risks in sediments. These findings emphasize that the continuous multi-source input and the resulting pseudo-persistence of PPCPs in sediments warrant prioritized continuous attention and targeted pollution control strategies in the Taipu River Basin. Full article

Monocular depth estimation for ${360}^{\circ }$ omnidirectional images is essential for immersive scene understanding and 3D reconstruction but remains challenging due to the non-uniform geometric distortions introduced by equirectangular projection (ERP). In particular, ERP suffers from latitude-dependent sampling bias, where the effective receptive field of standard convolution varies with the spherical latitude, leading to inconsistent feature representation and degraded depth prediction in high-distortion regions. To address this problem, this paper proposes a distortion-aware bi-projection fusion framework that integrates the global contextual continuity of ERP with the locally distortion-reduced geometric representation of cube map projection (CMP). The core component of the proposed framework is a Multi-scale Coordinate-Transformer Fusion (MCTF) module, which combines convolutional feature mixing, Transformer-based global context modeling, and Coordinate Attention-based spatial recalibration. By explicitly encoding vertical coordinate information into the fusion process, MCTF adaptively recalibrates feature responses according to latitude-dependent distortion patterns. Extensive experiments on the 3D60, Matterport3D, and Stanford2D3D benchmarks demonstrate that the proposed method consistently outperforms state-of-the-art omnidirectional depth estimation approaches. On the 3D60 dataset, our method reduces RMSE by 21.2% compared with the ERP baseline and achieves a 30% RMSE reduction in polar regions, where ERP distortion is most severe. These results validate the effectiveness of coordinate-aware feature calibration for robust ${360}^{\circ }$ monocular depth estimation. Full article

In high-pier bridge construction, the quality and accuracy of embedded rebar placement are critical to ensuring structural safety and durability. However, conventional manual inspection methods are inefficient, subjective and pose significant safety risks in high-altitude operations. These methods are unable to comprehensively inspect all pier columns on a daily basis, and frequently result in delays in acceptance that necessitate rework. In order to address these challenges, the current study proposes a smart vision-based inspection framework for the automatic and high-precision quality assessment of rebar under long-distance imaging conditions. This approach allows quality inspectors to remotely predict and evaluate the embedment quality of rebars from a safe distance. Notably, this work introduces a novel dual-source coordinate fusion mechanism that integrates improved instance segmentation with corner detection for global-to-local precision enhancement, representing an original contribution to rebar placement inspection in complex high-pier scenarios. The framework integrates an improved YOLOv8-CD segmentation model and a corner detection algorithm through a dual-source coordinate fusion mechanism, achieving an integration of global rebar detection and local feature enhancement. The YOLOv8-CD model, when optimised, features the Convolutional Block Attention Module (CBAM) integrated into the backbone, with the objective of enhancing recognition accuracy for small targets. Additionally, a Dilation-Wise Residual (DWR) module has been inserted before the neck C2f layer for the purpose of strengthening multi-scale feature extraction. The process of perspective correction and pixel-to-actual-length conversion coefficienting is performed in order to achieve a millimetre-level measurement of the rebar spacing and diameter. Empirical validation through real high-pier construction scenes demonstrates that the proposed framework attains a detection accuracy of 98.82%, surpassing conventional YOLO-based and single-source methodologies. The experimental results demonstrate that this framework is able to detect objects at longer distances, and to maintain its performance when the target is at a greater distance than that which was used for training. The proposed approach is expected to provide an efficient, safe, and quantitative solution for intelligent bridge construction quality monitoring, offering valuable insights for the future development of smart construction and structural health inspection systems. Full article

Sea fog is a common maritime meteorological hazard that poses a serious threat to maritime traffic safety and coastal economic activities. Existing sea fog detection models still struggle to effectively capture both global structural information and local textural details, while spectral confusion between clouds and fog further limits their performance. To address these challenges, a daytime sea fog detection model, HA-ENASnet, is developed using data from China’s Fengyun-4A (FY-4A) satellite and compared with six representative semantic segmentation models. Among the seven models, HA-ENASnet achieves the best performance, followed by SegMamba, scSE_LinkNet, SegMAN, ECA_TransUnet, DeepLabv3+, and UNet++. In terms of Intersection over Union (IoU), HA-ENASnet outperforms SegMamba and UNet++ by 3.08 and 17.26 percentage points, respectively. Ablation experiments for this model demonstrate that the introduction of a collaborative mechanism combining local and global linear attention effectively enhances the model’s perception capabilities for sea fog regions. Furthermore, the combination of Adaptive Architecture Search (NAS) and Efficient Channel Attention (ECA) facilitates the adaptive fusion of spectral and spatial multiscale features, thereby improving the model’s ability to distinguish between cloud and fog in spectrally confounded regions. Additionally, the proposed model is validated on Himawari-8 satellite data, exhibiting satisfactory detection performance and generalization capabilities. This provides methodological support for sea fog remote sensing monitoring using the FY satellite series. Full article

Fusing infrared and visible images can effectively compensate for the inherent limitations of each modality in different scenes, resulting in fused images that contain richer information. However, existing methods often struggle to balance global dependency modeling with local detail preservation and to effectively coordinate heterogeneous local and global features during fusion. To address these issues, this paper proposes a Significance-Preserving Progressive Fusion Network (SiPFusion). First, a progressive feature extraction framework was designed, which hierarchically extracts multi-scale local features using CNNs and then models long-range dependencies across scales via a Transformer-based global module. To adaptively integrate local-global complementary features, a significance-preserving fusion module was designed to obtain significance attention maps with a spatial selection mechanism, enabling dynamic fusion of multi-source features. Furthermore, we propose a significance similarity loss function that leverages intermediate feature guidance to enhance structural consistency and preserve salient-region information in the fused image. Extensive experiments on the MSRS, RoadScene, and TNO datasets demonstrate that SiPFusion achieves competitive visual quality and strong overall quantitative performance against 15 state-of-the-art fusion methods, obtaining leading results on most evaluated metrics. Full article

Power transmission line monitoring is critical for grid stability. However, existing object detection methods remain challenged by small-scale fittings in aerial transmission-line images because of feature sparsity, complex backgrounds, and computational constraints. In this paper, a YOLO-based dynamic adaptive sparse hierarchical network (DASH-YOLO) is proposed to address these issues. DASH-YOLO integrates dynamic adaptive computation with hierarchical attention to improve small-fitting feature representation while reducing redundant computation. Specifically, a Dynamic Sparse Hyper Graph (DSHG) module is introduced to regulate feature interaction density according to scene complexity, allowing relational modeling to focus on informative feature nodes. In addition, a Hierarchical Attention (HA) module adopts a coarse-to-fine refinement strategy to enhance global semantic awareness and local spatial precision. Experiments on a real-world transmission-line fitting dataset show that DASH-YOLO achieves 90.37% mAP@0.5 and 74.74% mAP@0.5:0.95, improving over YOLOv8x by 4.63 and 4.62 percentage points, respectively, while also reaching 90.80% recall, 85.49% precision, and 62.99 FPS under the same validation protocol. Full article

Gas concentration in coal mine return-air corners is affected by ventilation, mining disturbance and gas drainage conditions, and it shows strong nonstationarity, local fluctuation and dynamic multi-point correlations. To improve frequency information separation, monitoring point relationship modeling, and short-term prediction accuracy, a variational mode decomposition (VMD)-based dual-branch spatiotemporal graph method is proposed. Gas concentrations from four key monitoring points are used as inputs, and the return-air corner gas concentration is taken as the output. First, the raw series are decomposed by VMD and reconstructed into low- and high-frequency components. Then, two branches are built for different frequency components. The low-frequency branch combines adaptive graph learning, graph convolution and gated recurrent units to extract global variation features, while the high-frequency branch combines graph attention and gated recurrent units to capture local disturbance features. Finally, a feature-fusion module generates multi-step predictions, and a lightweight short-term warning strategy is developed based on the predicted values. The proposed model achieves MAE, RMSE and R² values of 0.0338, 0.0471 and 0.9499 in one-step prediction, respectively, and outperforms GRU, LSTM, GCN-GRU, GAT-GRU, VMD-GRU, Informer and STGCN under three-step and six-step conditions. Cross-dataset validation and inference time analysis indicate good adaptability and online prediction potential. Full article

Short-term power load forecasting is fundamental to the secure operation and optimal dispatch of modern power systems. This study proposes an Improved Pied Kingfisher Optimization–Temporal Convolutional Network–Bidirectional Gated Recurrent Unit (IPKO-TCN-BiGRU) model to address the challenges of strong non-stationarity, high randomness, and multi-factor coupling in load time series. The model employs a multi-scale TCN for simultaneous extraction of local and global temporal features, a BiGRU enhanced with an Improved Self-Attention (ISA) mechanism for bidirectional dependency modeling, and an Autoregressive (AR) module combined with an election mechanism to jointly capture linear and nonlinear load components. The Improved Pied Kingfisher Optimization (IPKO) algorithm—incorporating SPM chaotic initialization, a planetary optimization strategy, and adaptive t-distribution perturbation—is applied to globally optimize key hyperparameters, demonstrating superior convergence accuracy and global search capability over the original PKO and other benchmark optimizers. To ensure evaluation integrity, dataset splitting precedes all normalization operations, with StandardScaler fitted exclusively on the training set and applied to the test set without leakage. Validation is conducted on two benchmark datasets: a U.S. residential electricity load dataset (hourly, 2012, 13-dimensional features including HVAC and lighting systems) and a China Electrical Engineering Mathematical Modeling Competition dataset (15 min intervals, three years, enriched with five meteorological variables). The U.S. dataset exhibits a clear annual double-peak seasonal pattern, while the Chinese dataset shows strong intraday fluctuations significantly coupled with temperature and humidity, both posing substantial forecasting challenges. On the U.S. dataset, the proposed model achieves MAE = 0.0190 kW, RMSE = 0.0301 kW, MAPE = 1.7673%, and R² = 0.9947; on the China dataset, MAE = 79.8125 MW, RMSE = 109.4154 MW, MAPE = 1.1124%, and R² = 0.9955. The proposed model consistently outperforms six mainstream baseline models—including Transformer, Autoformer, and FEDformer—reducing RMSE by up to 34.4% and 18.9% on the two datasets, respectively, while maintaining a compact architecture of 15.2 MB and 74.6–78.9 MFLOPs. Ablation experiments confirm the significant and synergistic contribution of each module, and the direct comparison between PKO-TCN-BiGRU and IPKO-TCN-BiGRU validates that the algorithmic improvements translate into measurable forecasting gains beyond benchmark function optimization. The proposed model is most suitable for ultra-short-term to short-term single-step-ahead forecasting within a horizon of 15 min to 24 h, with an inference latency of 2.3–2.7 ms per sample, fully meeting the real-time requirements of practical power dispatching systems. Full article

X-ray coronary angiography remains the gold standard for diagnosing coronary heart disease. However, accurate segmentation is challenged by the subtlety of fine vascular features, topological discontinuities, and blurred boundaries in these images. Existing methods often struggle to capture both the long-range global topology and the fine local details required for robust segmentation. To address these issues, we propose BRNet, a framework that integrates a dual-backbone collaborative mechanism with multi-scale fusion and dynamic detail reconstruction. Our approach first employs a Vascular Local Detail Attention Module that combines ResNet18’s local perception with BiFormer’s global modeling, using SimAM parameter-free attention to suppress background noise. We then design a Vascular Structure-Coherent Progressive Fusion Module, which uses a top-down pyramid semantic flow to ensure topological coherence across different vascular scales. Finally, a Vascular Enhancement Dynamic Upsampling Module replaces traditional interpolation with content-aware CARAFE operators to achieve high-resolution reconstruction of blurred edges. Experiments on the public ARCADE dataset and the constructed heterogeneous benchmark ZZ-CAHDS show that BRNet achieves superior performance, attaining IoU scores of 0.6305 and 0.6774, and clDice coefficients of 0.7655 and 0.8355, respectively, and achieving a good balance between segmentation accuracy and computational efficiency. These results highlight BRNet’s effectiveness on retrospective segmentation benchmarks, demonstrating its capability for computer-assisted coronary artery segmentation. Full article

Accurate semantic segmentation of slender curvilinear structures, such as tree branches, in complex natural scenes remains challenging. The main difficulties arise from frequent occlusions, ambiguous boundaries, and limited edge computing resources. To address these issues, we propose LDS-Net, a lightweight dual-branch network designed for thin and continuous branch structures. The Detail-Aware Branch uses the Dynamic Snake Convolution (DSConv) to model irregular local geometry, while the Context-Aware Branch uses a Spatial Efficient Separable Pyramid module (SESP) to capture multi-scale context. To improve segmentation under occlusion and boundary ambiguity, we further integrate a Global Topology Transformer module (GTT) and a boundary guidance mechanism (BG). These features are fused via a Pixel-Wise Attention Fusion module (PAF) and optimized using a multi-head compound loss. Experiments on USTD and N-ABSD show that LDS-Net outperforms six representative networks. It also requires only 14.19G FLOPs, a 24% reduction compared with PIDNet-s. These results suggest that LDS-Net has potential for future deployment on resource-constrained agricultural and ecological monitoring platforms. Full article

Phenylketonuria (PKU) is an autosomal recessive inborn error of phenylalanine (Phe) metabolism caused by pathogenic variants in the phenylalanine hydroxylase (PAH) gene, resulting in toxic phenylalanine accumulation that, if untreated, causes profound intellectual disability and neurodevelopmental impairment. PKU is especially significant in the Kingdom of Saudi Arabia (KSA), where high consanguinity rates substantially elevate disease prevalence relative to Western populations, and the country’s expanding newborn screening programs have highlighted PKU as a persistent public health concern. This review provides a translational synthesis of the Saudi PKU literature, covering epidemiology, molecular pathophysiology, genetics, clinical presentation, diagnosis, treatment, prognosis, and future directions. We summarize global and regional incidence data and show that Saudi Arabia, driven by consanguinity, is among the countries with the highest reported PKU burden worldwide. We review the metabolic basis of phenylalanine neurotoxicity and the allelic heterogeneity of the PAH gene, with attention to variants enriched in Saudi and Arab cohorts, including the founder allele p.R252W. Diagnostic pathways anchored in newborn screening and tandem mass spectrometry are discussed alongside neurodevelopmental outcomes and gaps in Saudi PKU surveillance. We outline established and emerging therapies, including dietary management, sapropterin (BH4), pegvaliase, large neutral amino acids, and investigational gene and mRNA therapies. Throughout, we identify where genuine Saudi-specific evidence exists and where general PKU knowledge is extrapolated to the Saudi context because of limited local data, most notably the absence of a national PKU registry. This review is intended to serve as a translational reference for clinicians, metabolic dietitians, geneticists, and policymakers engaged in PKU care in Saudi Arabia and the wider Arab region. Full article

Accurate footprint image segmentation is challenging in forensic applications because fine anatomical structures, weak boundaries, and background interference can degrade segmentation performance. This study presents a task-oriented generative adversarial network (GAN)-based framework for forensic footprint image segmentation. Channel Prior Convolutional Attention (CPCA) modules are integrated into the decoder stages of the generator to recalibrate fused encoder–decoder features and preserve fine details in the toe, arch, and heel regions. In addition, a dual-branch discriminator processes image–mask pairs at the original and downsampled scales, providing complementary constraints on local boundary details and global footprint morphology. The framework is trained with a least-squares adversarial loss and a binary cross-entropy (BCE)–Dice segmentation loss. Experiments on the self-collected aFoot_2025 dataset show that the proposed framework achieves an IoU of 0.9448 and a Dice coefficient of 0.9713, outperforming the evaluated baseline and attention-based alternatives. Under the evaluated synthetic Gaussian-noise settings, the proposed method retained relatively stable segmentation performance. Furthermore, an exploratory footprint-based height-prediction analysis showed modestly lower prediction errors than the baseline GAN. These findings indicate that, under the controlled acquisition conditions of the aFoot_2025 dataset, CPCA-based feature calibration and dual-scale discrimination may improve segmentation-mask quality and provide a possible benefit for subsequent anthropometric analysis. Full article