Search Results (5,597)

As a well-known World Cultural Heritage Site, the Historic Centre of Macao’s historical buildings possess a wealth of decorative patterns. These patterns contain cultural esthetics, geographical environment, cultural traditions, and other elements from specific historical periods, deeply reflecting the evolution of religious rituals and political and economic systems throughout history. Through long-term research, this article constructs a dataset of 11,807 images of local decorative patterns of historical buildings in Macau, and proposes a fine-grained image classification method using the ConvNeXt-L model. The ConvNeXt-L model is an efficient convolutional neural network that has demonstrated excellent performance in image classification tasks in fields such as medicine and architecture. Its outstanding advantages lie in limited training samples, diverse image features, and complex scenes. The most typical advantage of this model is its structural integration of key design concepts from a Transformer, which significantly enhances the feature extraction and generalization ability of samples. In response to the objective reality that the decorative patterns of historical buildings in Macau have rich levels of detail and a limited number of functional building categories, ConvNeXt-L maximizes its ability to recognize and classify patterns while ensuring computational efficiency. This provides a more ideal technical path for the classification of small-sample complex images. This article constructs a deep learning system based on the PyTorch 1.11 framework and compares ResNet50, EfficientNet-B7, ViT-B/16, Swin-B, RegNet-Y-16GF, and ConvNeXt series models. The results indicate a positive correlation between model performance and structural complexity, with ConvNeXt-L being the most ideal in terms of accuracy in decorative pattern classification, due to its fusion of convolution and attention mechanisms. This study not only provides a multidimensional exploration for the protection and revitalization of Macao’s historical and cultural heritage and enriches theoretical support and practical foundations but also provides new research paths and methodological support for artificial intelligence technology to assist in the planning and decision-making of historical urban areas. Full article

This study develops an innovative unmanned aerial vehicle (UAV)-based intelligent system for winter wheat yield prediction, addressing the inefficiencies of traditional manual counting methods (with approximately 15% error rate) and enabling quantitative analysis of water–fertilizer interactions. By integrating an enhanced Faster Region-Based Convolutional Neural Network (Faster R-CNN) architecture with multi-source data fusion and machine learning, the system significantly improves both spike detection accuracy and yield forecasting performance. Field experiments during the 2022–2023 growing season captured high-resolution multispectral imagery for varied irrigation regimes and fertilization treatments. The optimized detection model incorporates ResNet-50 as the backbone feature extraction network, with residual connections and channel attention mechanisms, achieving a mean average precision (mAP) of 91.2% (calculated at IoU threshold 0.5) and 88.72% recall while reducing computational complexity. The model outperformed YOLOv8 by a statistically significant 2.1% margin (p < 0.05). Using model-generated spike counts as input, the random forest (RF) model regressor demonstrated superior yield prediction performance (R² = 0.82, RMSE = 324.42 kg·ha⁻¹), exceeding the Partial Least Squares Regression (PLSR) (R² +46%, RMSE-44.3%), Least Squares Support Vector Machine (LSSVM) (R² + 32.3%, RMSE-32.4%), Support Vector Regression (SVR) (R² + 30.2%, RMSE-29.6%), and Backpropagation (BP) Neural Network (R²+22.4%, RMSE-24.4%) models. Analysis of different water–fertilizer treatments revealed that while organic fertilizer under full irrigation (750 m³ ha⁻¹) conditions achieved maximum yield benefit (13,679.26 CNY·ha⁻¹), it showed relatively low water productivity (WP = 7.43 kg·m⁻³). Conversely, under deficit irrigation (450 m³ ha⁻¹) conditions, the 3:7 organic/inorganic fertilizer treatment achieved optimal WP (11.65 kg m⁻³) and WUE (20.16 kg∙ha⁻¹∙mm⁻¹) while increasing yield benefit by 25.46% compared to organic fertilizer alone. This research establishes an integrated technical framework for high-throughput spike monitoring and yield estimation, providing actionable insights for synergistic water–fertilizer management strategies in sustainable precision agriculture. Full article

This study investigates the hydrodynamic loads of “Ningde No. 1” offshore aquaculture under current-only conditions using a fluid–structure interaction (FSI) approach with the computational fluid dynamics (CFD) solver OpenFOAM. A porous-media-based model is applied to simulate net-induced drag, while the rigid framework is resolved using a large eddy simulation (LES) turbulence model. A comprehensive set of 350 CFD simulations is performed, with varying flow velocities, flow directions, draft depths, and existence of nets. The results reveal that the load on this fishing facility in the streamwise direction (Fx) increases monotonically with flow velocity, direction, and draft. The lateral (Fy) and vertical (Fz) loads exhibit non-linear trends, peaking at a specific flow direction (approximately 60°) and draft levels (around 11.5 m). The fishing nets substantially increase the streamwise load by up to 80%, while their influence on the lateral forces is dependent on submergence depth. To efficiently predict hydrodynamic loads without performing additional and lengthy CFD simulations, a physics-informed neural network (PINN) is trained using the simulated data. The PINN model is found able to accurately reproduce the hydrodynamic force across a wide range of current conditions, offering a practical and interpretable surrogate approach for structural design optimization and mooring system development in offshore aquaculture industry. Full article

Tennis serves heavily impact match outcomes, yet analysis by coaches is limited by human vision. The design of an automated tennis serve analysis system could facilitate enhanced performance analysis. As serve location and serve success are directly correlated, predicting the outcome of a serve could provide vital information for performance analysis. This article proposes a tennis serve analysis system powered by Machine Learning, which classifies the outcome of serves as “in”, “out” or “net”, and predicts the coordinate outcome of successful serves. Additionally, this work details the collection of three-dimensional spatio-temporal data on tennis serves, using marker-based optoelectronic motion capture. The classification uses a Stacked Bidirectional Long Short-Term Memory architecture, whilst a 3D Convolutional Neural Network architecture is harnessed for serve coordinate prediction. The proposed method achieves 89% accuracy for tennis serve classification, outperforming the current state-of-the-art whilst performing finer-grain classification. The results achieve an accuracy of 63% in predicting the serve coordinates, with a mean absolute error of 0.59 and a root mean squared error of 0.68, exceeding the current state-of-the-art with a new method. The system contributes towards the long-term goal of designing a non-invasive tennis serve analysis system that functions in training and match conditions. Full article

Over the years, researchers have developed several speed estimation techniques using wearable inertial measurement units (IMUs). In this paper, we introduce a medium-scale dataset, containing measurements of walking/running at speeds ranging from 4.0 km/h (1.11 m/s) to 9.5 km/h (2.64 m/s) in increments of 0.5 km/h (0.14 m/s) from 33 healthy subjects wearing IMUs. We name it the “Speed” dataset. In summary, we present accelerometer and gyroscope data from 12 speeds and 22 subject-independent sets with the full range of 12 speeds. The data in each set consists of overlapping sections of 250 time samples (corresponding to 2.5 s, sampled at 100 Hz), and six dimensions (corresponding to the three axes of the accelerometer and three axes of the gyroscope). Each speed set contains 1775 examples. We benchmark the existing approaches used in the literature for the purpose of speed estimation on this dataset. These include support vector regression, Gaussian Process Regression, and shallow neural networks. We then design a deep Convolutional Neural Network (CNN), SpeedNet, for baseline results. The proposed SpeedNet yields an average Root Mean Square Error (RMSE) of 0.4819 km/h (0.13 m/s), following a subject-independent approach. Then, the SpeedNet obtained from the subject-independent approach are adapted using a portion of subject-specific data. The average RMSE for the remainder of the data for all subjects then drops down to 0.1747 km/h (0.05 m/s). The suggested SpeedNet yields a lower RMSE in comparison to the other approaches. In addition, we also compare the proposed method to others in terms of the average testing time, to give an idea of computational complexity. The proposed SpeedNet, despite being more accurate, yields real-time performance. Full article

Accurate automatic sleep stage classification from single-channel EEG remains challenging due to the need for effective extraction of multiscale neurophysiological features and modeling of long-range temporal dependencies. This study aims to address these limitations by developing an efficient and compact deep learning architecture tailored for wearable and edge device applications. We propose MultiScaleSleepNet, a hybrid convolutional neural network–bidirectional long short-term memory–transformer architecture that extracts multiscale temporal and spectral features through parallel convolutional branches, followed by sequential modeling using a BiLSTM memory network and transformer-based attention mechanisms. The model obtained an accuracy, macro-averaged F1 score, and kappa coefficient of 88.6%, 0.833, and 0.84 on the Sleep-EDF dataset; 85.6%, 0.811, and 0.80 on the Sleep-EDF Expanded dataset; and 84.6%, 0.745, and 0.79 on the SHHS dataset. Ablation studies indicate that attention mechanisms and spectral fusion consistently improve performance, with the most notable gains observed for stages N1, N3, and rapid eye movement. MultiScaleSleepNet demonstrates competitive performance across multiple benchmark datasets while maintaining a compact size of 1.9 million parameters, suggesting robustness to variations in dataset size and class distribution. The study supports the feasibility of real-time, accurate sleep staging from single-channel EEG using parameter-efficient deep models suitable for portable systems. Full article

Cracks are the initial manifestation of various diseases on highways. Preventive maintenance of cracks can delay the degree of pavement damage and effectively extend the service life of highways. However, existing crack detection methods have poor performance in identifying small cracks and are unable to calculate crack width, leading to unsatisfactory preventive maintenance results. This article proposes an integrated method for crack detection, segmentation, and width calculation based on digital image processing technology. Firstly, based on convolutional neural network, a optimized crack detection network called CFSSE is proposed by fusing the fast spatial pyramid pooling structure with the squeeze-and-excitation attention mechanism, with an average detection accuracy of 97.10%, average recall rate of 98.00%, and average detection precision at 0.5 threshold of 98.90%; it outperforms the YOLOv5-mobileone network and YOLOv5-s network. Secondly, based on the U-Net network, an optimized crack segmentation network called CBU_Net is proposed by using the CNN-block structure in the encoder module and a bicubic interpolation algorithm in the decoder module, with an average segmentation accuracy of 99.10%, average intersection over union of 88.62%, and average pixel accuracy of 93.56%; it outperforms the U_Net network, DeepLab v3+ network, and optimized DeepLab v3 network. Finally, a laser spot center positioning method based on information entropy combination is proposed to provide an accurate benchmark for crack width calculation based on parallel lasers, with an average error in crack width calculation of less than 2.56%. Full article

This study presents a robust and extensible hybrid classification framework for accurately detecting diseases in citrus leaves by integrating transfer learning-based deep learning models with classical machine learning techniques. Features were extracted using advanced pretrained architectures—DenseNet201, ResNet50, MobileNetV2, and EfficientNet-B0—and refined via the minimum redundancy maximum relevance (mRMR) method to reduce redundancy while maximizing discriminative power. These features were classified using support vector machines (SVMs), ensemble bagged trees, k-nearest neighbors (kNNs), and neural networks under stratified 10-fold cross-validation. On the lemon dataset, the best configuration (DenseNet201 + SVM) achieved 94.1 ± 4.9% accuracy, 93.2 ± 5.7% F1 score, and a balanced accuracy of 93.4 ± 6.0%, demonstrating strong and stable performance. To assess external generalization, the same pipeline was applied to mango and pomegranate leaves, achieving 100.0 ± 0.0% and 98.7 ± 1.5% accuracy, respectively—confirming the model’s robustness across citrus and non-citrus domains. Beyond accuracy, lightweight models such as EfficientNet-B0 and MobileNetV2 provided significantly higher throughput and lower latency, underscoring their suitability for real-time agricultural applications. These findings highlight the importance of combining deep representations with efficient classical classifiers for precision agriculture, offering both high diagnostic accuracy and practical deployability in field conditions. Full article

Accurate and timely prediction of diseases in water-intensive crops is critical for sustainable agriculture and food security. AI-based crop disease management tools are essential for an optimized approach, as they offer significant potential for enhancing yield and sustainability. This study centers on maize, training deep learning models on UAV imagery and satellite remote-sensing data to detect and predict disease. The performance of multiple convolutional neural networks, such as ResNet-50, DenseNet-121, etc., is evaluated by their ability to classify maize diseases such as Northern Leaf Blight, Gray Leaf Spot, Common Rust, and Blight using UAV drone data. Remotely sensed MODIS satellite data was used to generate spatial severity maps over a uniform grid by implementing time-series modeling. Furthermore, reinforcement learning techniques were used to identify hotspots and prioritize the next locations for inspection by analyzing spatial and temporal patterns, identifying critical factors that affect disease progression, and enabling better decision-making. The integrated pipeline automates data ingestion and delivers farm-level condition views without manual uploads. The combination of multiple remotely sensed data sources leads to an efficient and scalable solution for early disease detection. Full article

Under the context of “Energy Conservation and Emission Reduction,” low-sulfur fuel has become widely adopted in maritime operations, posing significant challenges to cylinder lubrication systems. Traditional oil injection strategies, heavily reliant on manual experience, suffer from instability and high costs. To address this, a lightweight image retrieval model for cylinder lubrication is proposed, leveraging deep learning and computer vision to support oiling decisions based on visual features. The model comprises three components: a backbone network, a feature enhancement module, and a similarity retrieval module. Specifically, EfficientNetB0 serves as the backbone for efficient feature extraction under low computational overhead. MobileViT Blocks are integrated to combine local feature perception of Convolutional Neural Networks (CNNs) with the global modeling capacity of Transformers. To further improve receptive field and multi-scale representation, Receptive Field Blocks (RFB) are introduced between the components. Additionally, the Convolutional Block Attention Module (CBAM) attention mechanism enhances focus on salient regions, improving feature discrimination. A high-quality image dataset was constructed using WINNING’s large bulk carriers under various sea conditions. The experimental results demonstrate that the EfficientNetB0 + RFB + MobileViT + CBAM model achieves excellent performance with minimal computational cost: 99.71% Precision, 99.69% Recall, and 99.70% F1-score—improvements of 11.81%, 15.36%, and 13.62%, respectively, over the baseline EfficientNetB0. With only a 0.3 GFLOP and 8.3 MB increase in model size, the approach balances accuracy and inference efficiency. The model also demonstrates good robustness and application stability in real-world ship testing, with potential for further adoption in the field of intelligent ship maintenance. Full article

As oil and gas exploration and development advance into deep and ultra-deep areas, kick accidents are becoming more frequent during drilling operations, posing a serious threat to construction safety. Traditional kick monitoring methods are limited in their multivariate coupling modeling. These models rely too heavily on single-feature weights, making them prone to misjudgment. Therefore, this paper proposes a drilling kick risk prediction method based on image modality. First, a sliding window mechanism is used to slice key drilling parameters in time series to extract multivariate data for continuous time periods. Second, data processing is performed to construct joint logging curve image samples. Then, classical CNN models such as VGG16 and ResNet are used to train and classify image samples; finally, the performance of the model on a number of indicators is evaluated and compared with different CNN and temporal neural network models. Finally, the model’s performance is evaluated across multiple metrics and compared with CNN and time series neural network models of different structures. Experimental results show that the image-based VGG16 model outperforms typical convolutional neural network models such as AlexNet, ResNet, and EfficientNet in overall performance, and significantly outperforms LSTM and GRU time series models in classification accuracy and comprehensive discriminative power. Compared to LSTM, the recall rate increased by 23.8% and the precision increased by 5.8%, demonstrating that its convolutional structure possesses stronger perception and discriminative capabilities in extracting local spatiotemporal features and recognizing patterns, enabling more accurate identification of kick risks. Furthermore, the pre-trained VGG16 model achieved an 8.69% improvement in accuracy compared to the custom VGG16 model, fully demonstrating the effectiveness and generalization advantages of transfer learning in small-sample engineering problems and providing feasibility support for model deployment and engineering applications. Full article

The monitoring and diagnostics of energy equipment aim to detect anomalies in time series data in order to support predictive maintenance and avoid unplanned shutdowns. Thus, the paper proposes a novel methodology that utilizes sequence-to-image transformation methods to feed Convolutional Neural Networks (CNNs) for diagnostic purposes. Multivariate time series taken from real gas turbines are transformed by using two methods. We study two CNN architectures, i.e., VGG-19 and SqueezeNet. The investigated anomaly is the spike fault. Spikes are implanted in field multivariate time series taken during normal operation of ten gas turbines and composed of twenty gas path measurements. Six fault scenarios are simulated. For each scenario, different combinations of fault parameters are considered. The main novel contribution of this study is the development of a comprehensive framework, which starts from time series transformation and ends up with a diagnostic response. The potential of CNNs for image recognition is applied to the gas path field measurements of a gas turbine. A hard-to-detect type of fault (i.e., random spikes of different magnitudes and frequencies of occurrence) was implanted in a seemingly real-world fashion. Since spike detection is highly challenging, the proposed framework has both scientific and industrial relevance. The extended and thorough analyses unequivocally prove that CNNs fed with images are remarkably more accurate than TCN models fed with raw time series data, with values higher than 93% if the number of implanted spikes is 10% of the total data and a gain in accuracy of up to 40% in the most realistic scenario. Full article

Tourism and digitalization have become increasingly interconnected, yet the complex, nonlinear relationships between technological adoption and tourism performance remain underexplored. This study aims to examine how enterprise software solutions influence tourism indicators across European countries. Using a triangulated methodological approach, we employed factor analysis to identify underlying dimensions, neural network modeling to detect nonlinear relationships, and hierarchical clustering to group countries based on digital and tourism profiles. The results consistently highlight CRM (Customer Relationship Management) as the most influential technological factor linked to both the net occupancy rate of beds and the number of nights spent at tourist accommodations. While AI (artificial intelligence) technologies currently have less impact, their importance is growing, as seen in emerging patterns. Cluster analysis further confirms that countries with higher CRM adoption tend to cluster together and show better tourism performance, indicating a clear connection between digital maturity and sector competitiveness. These findings emphasize the strategic importance of CRM as a transformative tool in hospitality and tourism management, while also recognizing the potential of AI to shape future trends. The study offers empirical support for tailored digital policies across European regions to promote inclusive and sustainable tourism growth. Full article

Sewage sludge, as a by-product of wastewater treatment, represents a significant cost factor in the operation of wastewater treatment plants and accounts for up to 50% of total costs. As sewage sludge still contains a high proportion of water after the basic treatment processes (thickening, stabilization and dewatering), sludge drying helps to reduce further treatment and disposal costs. Conventional drying methods are associated with high energy consumption, making solar drying a more cost-effective alternative. This paper analyzes the economic aspects of constructing a sewage sludge solar drying facility with the help of an expert system based on neural networks. The system considers a range of parameters (plant capacity, transport distance, transport and treatment costs, etc.) to assess the values of the investment as well as the operation and maintenance costs. The analysis was carried out using NeuralTools (Lumivero). Two main options for sludge disposal were investigated: treatment at a regional center (with the sub-options of own or outsourced transport) and handing over of sludge to another legal entity. In total, five neural network models were developed based on the input load (from 75 to 10,000 t/year and from 10,000 to 20,000 t/year) and transport method (own or outsourced transport), resulting in an analysis of over 670,000 scenarios. The key output variable was the net present value of costs over a 30-year period. The results demonstrated high model accuracy (error < 5%) and allowed a comparison of the profitability of constructing a sewage sludge solar drying facility with alternative methods of sludge disposal, in particular with the transport and disposal of the dewatered sludge. Full article

Single-image dehazing of high-voltage power transmission lines (HPTLs) using deep learning methods confronts two critical challenges: the non-homogeneous haze distribution in HPTL images and the unavailability of paired clear images for supervised training. To overcome these issues, this paper proposes a novel dehaze neural network, named FIF-RSCT-Net, that employs a hybrid supervised-to-unsupervised iterative learning approach according to the characteristic of HPTL single images. The FIF-RSCT-Net incorporates the Spatial–Channel Feature Intersection modules and Residual Separable Convolution Transformers to enhance the feature representation capability. Crucially, this novel architecture could learn more generalized dehazing knowledge that can be transferred from the original image domain to HPTL scenarios. In the dehazing knowledge transformation, an unsupervised iterative learning mechanism based on the Line Segment Detector is designed to optimize the restoration of power transmission lines. The effectiveness of FIF-RSCT-Net on the original image domain is demonstrated in the comparative experiments of the I-Haze, O-Haze, NH-Haze, and SOTS datasets. Our methodology achieves the best average PSNR of 24.647 dB and SSIM of 0.8512. And the qualitative evaluation of unsupervised iterative learning results shows that the missed line segments are exhibited during progressive training iterations. Full article