Search Results (1,580)

Product quantization (PQ) is a widely adopted technique for efficient approximate nearest neighbor (ANN) search in high-dimensional spaces, offering a favorable balance between accuracy and memory efficiency. However, standard PQ suffers from high online computational cost when the number of subspaces is high. To address this dilemma, we propose Pyramid Product Quantization (PPQ), a novel adaptive quantization framework that dynamically selects the most suitable number of subspaces for different segments of each data vector. This leads to a significant reduction in the number of addition operations required during approximate distance computation, significantly accelerating online search. Experimental results demonstrate that the proposed PPQ method effectively lowers the computational complexity of product quantization and its variants, without compromising retrieval accuracy. Full article

Traditional routing algorithms optimizing for distance or travel time are inadequate for electric vehicles (EVs), which require energy-aware planning considering battery constraints and charging infrastructure. This work presents an energy-optimal routing system for EVs that integrates personalized consumption modeling with real-time environmental data. The system employs a Long Short-Term Memory (LSTM) neural network to predict State-of-Charge (SoC) consumption from real-world driving data, learning directly from spatiotemporal features including velocity, temperature, road inclination, and traveled distance. Unlike physics-based models requiring difficult-to-obtain parameters, this approach captures nonlinear dependencies and temporal patterns in energy consumption. The routing framework integrates static map data, dynamic traffic conditions, weather information, and charging station locations into a weighted graph representation. Edge costs reflect predicted SoC drops, while node penalties account for traffic congestion and charging opportunities. An enhanced A* algorithm finds optimal routes minimizing energy consumption. Experimental validation on a Nissan Leaf shows that the proposed end-to-end SoC estimator significantly outperforms traditional approaches. The model achieves an RMSE of 36.83 and an $R^2$ of 0.9374, corresponding to a 59.91% reduction in error compared to physics-based formulas. Real-world testing on various routes further confirms its accuracy, with a Mean Absolute Error in the total route SoC estimation of 2%, improving upon the 3.5% observed for commercial solutions. Full article

In the pursuit of real-time object detection with constrained computational resources, the optimization of neural network architectures is paramount. We introduce novel sparsity induction methods within the YOLOv4-Tiny framework to significantly improve computational efficiency while maintaining high accuracy in pedestrian detection. We present three sparsification approaches: Homogeneous, Progressive, and Layer-Adaptive, each methodically reducing the model’s complexity without compromising its detection capability. Additionally, we refine the model’s output with a memory-efficient sliding window approach and a Bounding Box Sorting Algorithm, ensuring precise Intersection over Union (IoU) calculations. Our results demonstrate a substantial reduction in computational load by zeroing out over 50% of the weights with only a minimal 6% loss in IoU and 0.6% loss in F1-Score. Full article

Environmental monitoring systems generate large volumes of multivariate time series data from heterogeneous sensors, including those measuring soil, weather, and air quality parameters. However, sensor malfunctions and transmission failures frequently lead to missing values, compromising the performance of downstream analytical and predictive models. To address this challenge, this study presents a comprehensive and systematic evaluation of previously proposed hybrid architecture that interleaves Long Short-Term Memory (LSTM) layers with a Multi-Head Attention mechanism in a “sandwiched” setting (LSTM–Attention–LSTM) for robust multivariate data imputation in environmental IoT datasets. The first LSTM layer captures short-term temporal dependencies, the attention layer emphasises long-range relationships among correlated features, and the second LSTM layer re-integrates these enriched representations into a coherent temporal sequence. The model is evaluated using multiple environmental datasets of soil temperature, meteorological (precipitation, temperature, wind speed, humidity), and air quality data across missingness levels ranging from 10% to 90%. Performance is compared against baseline methods, including K-Nearest Neighbour (KNN) and Bidirectional Recurrent Imputation for Time Series (BRITS). Across all datasets, the Hybrid model consistently outperforms baseline methods, achieving MAE reductions exceeding 50% and reaching over 80% in several scenarios, along with RMSE reductions of up to approximately 85%, particularly under moderate to high missingness conditions. An ablation study further examines the contribution of each layer to overall model performance. Results demonstrate that the proposed Hybrid model achieves superior accuracy and robustness across datasets, confirming its effectiveness for environmental sensor data imputation under varying missing data conditions. Full article

The fermentation process of oolong tea is a critical step in shaping its quality and flavor profile. In this study, the fermentation degree of Anxi Tieguanyin oolong tea was assessed using image and hyperspectral features. Machine learning algorithms, including Support Vector Machine (SVM), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU), were employed to develop models based on both single-source features and multi-source fused features. First, color and texture features were extracted from RGB images and then processed through Pearson correlation-based feature selection and Principal Component Analysis (PCA) for dimensionality reduction. For the hyperspectral data, preprocessing was conducted using Normalization (Nor) and Standard Normal Variate (SNV), followed by feature selection and dimensionality reduction with Competitive Adaptive Reweighted Sampling (CARS), Successive Projections Algorithm (SPA), and PCA. We then performed mid-level fusion on the two feature sets and selected the most relevant features using L1 regularization for the final modeling stage. Finally, SHapley Additive exPlanations (SHAP) analysis was conducted on the optimal models to reveal key features from both hyperspectral bands and image data. The results indicated that models based on single features achieved test set accuracies of 68.06% to 87.50%, while models based on data fusion achieved 77.78% to 94.44%. Specifically, the Pearson+Nor-SPA+L1+SVM fusion model achieved the highest accuracy of 94.44%. This demonstrates that data feature fusion enables a more comprehensive characterization of the fermentation process, significantly improving model accuracy. SHAP analysis revealed that the hyperspectral bands at 967, 942, 814, 784, 781, 503, 413, and 416 nm, along with the image features H_σ and H, played the most crucial roles in distinguishing tea fermentation stages. These findings provide a scientific basis for assessing the fermentation degree of Tieguanyin oolong tea and support the development of intelligent detection systems. Full article

Marine oil spill detection using Synthetic Aperture Radar (SAR) is crucial but challenged by dynamic marine conditions, diverse spill scales, and limitations in existing algorithms regarding model size and real-time performance. To address these challenges, we propose LSFE-YOLO, a YOLOv8s-optimized (You Only Look Once version 8) lightweight model with an original, domain-tailored synergistic integration of FasterNet, GN-LSC Head (GroupNorm Lightweight Shared Convolution Head), and C2f_MBE (C2f Mobile Bottleneck Enhanced). FasterNet serves as the backbone (25% neck width reduction), leveraging partial convolution (PConv) to minimize memory access and redundant computations—overcoming traditional lightweight backbones’ high memory overhead—laying the foundation for real-time deployment while preserving feature extraction. The proposed GN-LSC Head replaces YOLOv8’s decoupled head: its shared convolutions reduce parameter redundancy by approximately 40%, and GroupNorm (Group Normalization) ensures stable accuracy under edge computing’s small-batch constraints, outperforming BatchNorm (Batch Normalization) in resource-limited scenarios. The C2f_MBE module integrates EffectiveSE (Effective Squeeze and Excitation)-optimized MBConv (Mobile Inverted Bottleneck Convolution) into C2f: MBConv’s inverted-residual design enhances multi-scale feature capture, while lightweight EffectiveSE strengthens discriminative oil spill features without extra computation, addressing the original C2f’s scale variability insufficiency. Additionally, an SE (Squeeze and Excitation) attention mechanism embedded upstream of SPPF (Spatial Pyramid Pooling Fast) suppresses background interference (e.g., waves, biological oil films), synergizing with FasterNet and C2f_MBE to form a cascaded feature optimization pipeline that refines representations throughout the model. Experimental results show that LSFE-YOLO improves mAP (mean Average Precision) by 1.3% and F1 score by 1.7% over YOLOv8s, while achieving substantial reductions in model size (81.9%), parameter count (82.9%), and computational cost (84.2%), alongside a 20 FPS (Frames Per Second) increase in detection speed. LSFE-YOLO offers an efficient and effective solution for real-time marine oil spill detection. Full article

The canopy temperature of winter jujube serves as a direct indicator of plant water status and transpiration efficiency, making its accurate prediction a critical prerequisite for effective water management and optimized growth conditions in greenhouse environments. This study developed a data-driven model to forecast canopy temperature. The model serially integrates a Long Short-Term Memory (LSTM) network and a Random Forest (RF) algorithm, leveraging their complementary strengths in capturing temporal dependencies and robust nonlinear fitting. A three-stage framework comprising temporal feature extraction, multi-source feature fusion, and direct prediction was implemented to enable reliable nowcasting. Data acquisition and preprocessing were tailored to the greenhouse environment, involving multi-sensor data and thermal imagery processed with Robust Principal Component Analysis (RPCA) for dimensionality reduction. Key environmental variables were selected through Spearman correlation analysis. Experimental results demonstrated that the proposed LSTM–RF model achieved superior performance, with a determination coefficient (R²) of 0.974, mean absolute error (MAE) of 0.844 °C, and root mean square error (RMSE) of 1.155 °C, outperforming benchmark models including standalone LSTM, RF, Transformer, and TimesNet. SHAP (SHapley Additive exPlanations)-based interpretability analysis further quantified the influence of key factors, including the “thermodynamic state of air” driver group and latent temporal features, offering actionable insights for irrigation management. The model establishes a reliable, interpretable foundation for real-time water stress monitoring and precision irrigation control in protected winter jujube production systems. Full article

This work investigates the effectiveness of block transform coding (BTC) as a lightweight, training-free quantization strategy for compressing the weights of pretrained deep neural networks. The proposed method applies a rule-based block transform with variance and root mean square error (RMSE)-driven stopping criteria, enabling substantial reductions in bit precision while preserving the statistical structure of convolutional and fully connected layer weights. Unlike uniform 8-bit quantization, BTC dynamically adjusts bit usage across layers and achieves significantly lower distortion for the same compression budget. We evaluate BTC across many pretrained architectures and tabular benchmarks. Experimental results show that BTC consistently reduces storage to 4–7.7 bits per weight while maintaining accuracy within 2–3% of the 32-bit floating point (FP32) baseline. To further assess scalability and baseline strength, BTC is additionally evaluated on large-scale ImageNet models and compared against a calibrated percentile-based uniform post-training quantization method. The results show that BTC achieves a substantially lower effective bit-width while incurring only a modest accuracy reduction relative to calibration-aware 8-bit quantization, highlighting a favorable compression–accuracy trade-off. BTC also exhibits stable behavior across successive post-training quantization (PTQ) configurations, low quantization noise, and smooth RMSE trends, outperforming naïve uniform quantization under aggressive compression. These findings confirm that BTC provides a scalable, architecture-agnostic, and training-free quantization mechanism suitable for deployment in memory- and computing-constrained environments. Full article

To develop a cost-effective shape memory alloy fiber-reinforced engineered cementitious composite (SMAF-ECC) with excellent mechanical properties, polypropylene (PP) fibers were used to partially replace polyvinyl alcohol (PVA) fibers to prepare the ECC matrix, and superelastic shape memory alloy fibers (SMAFs) were incorporated to fabricate a novel SMAF-ECC. Uniaxial tensile tests were systematically performed to characterize the tensile mechanical properties of the composites, focusing on the effects of SMAF volume content and diameter. The results indicate that the optimal base ECC mix proportion is 0.8 vol.% PP fibers and 1.2 vol.% PVA fibers, achieving an ultimate tensile strain of 4.88% (only a 4.69% reduction compared to pure PVA-ECC) while significantly reducing material cost without sacrificing superior ductility. SMAF volume content and diameter notably influence the tensile performance of SMAF-ECC, with the specimen containing 0.2 mm diameter SMAFs at 0.2 vol.% exhibiting the best performance: initial cracking stress, ultimate tensile stress, and ultimate tensile strain are enhanced by 16.79%, 20.85%, and 2.87%, respectively, compared to pure ECC. This study provides a theoretical basis and parametric guidance for the engineering popularization and application of cost-effective SMAF-ECCs. Full article

The exponential growth of data-driven applications has intensified performance demands on database systems, where cache management represents a critical bottleneck. Traditional cache replacement policies such as Least Recently Used (LRU) and Least Frequently Used (LFU) rely on simple heuristics that fail to capture complex temporal and frequency patterns in modern workloads. This research presents a modular machine learning-enhanced cache management framework that leverages pattern recognition to optimize database performance through intelligent replacement decisions. Our approach integrates multiple machine learning models—Random Forest classifiers, Long Short-Term Memory (LSTM) networks, Support Vector Machines (SVM), and Gradient Boosting methods—within a modular architecture enabling seamless integration with existing database systems. The framework incorporates sophisticated feature engineering pipelines extracting temporal, frequency, and contextual characteristics from query access patterns. Comprehensive experimental evaluation across synthetic workloads, real-world production datasets, and standard benchmarks (TPC-C, TPC-H, YCSB, and LinkBench) demonstrates consistent performance improvements. Machine learning-enhanced approaches achieve $8.4\%$ to $19.2\%$ improvement in cache hit rates, $15.3\%$ to $28.7\%$ reduction in query latency, and $18.9\%$ to $31.4\%$ increase in system throughput compared to traditional policies and advanced adaptive methods including ARC, LIRS, Clock-Pro, TinyLFU, and LECAR. Random Forest emerges as the most practical solution, providing $18.7\%$ performance improvement with only $3.1\%$ computational overhead. Case study analysis across e-commerce, financial services, and content management applications demonstrates measurable business impact, including $8.3\%$ conversion rate improvements and USD 127,000 annual revenue increases. Statistical validation ($p<0.001$, Cohen’s $d>0.8$) confirms both statistical and practical significance. Full article

Background/Objectives: Post-traumatic epileptogenesis is a frequent and clinically relevant consequence of traumatic brain injury (TBI), often contributing to worsened neurological and functional outcomes. In patients experiencing early post-injury seizures, rehabilitative strategies that support recovery while considering increased epileptogenic risk are needed. This case study explores the potential benefits of combining neurofeedback (NFB) with motor therapy on cognitive and motor recovery. Methods: A patient hospitalized for severe TBI who experienced an acute symptomatic seizure in the early post-injury phase underwent baseline quantitative EEG (qEEG), neuromotor, functional, and neuropsychological assessments. The patient then completed a three-week rehabilitation program (five days/week) including 30 sensorimotor rhythm (SMR) NFB sessions (35 min each) combined with daily one-hour motor therapy. qEEG and clinical assessments were repeated post-intervention and at 6-month follow-up. Results: Post-intervention qEEG showed significant reductions in Delta and Theta power, reflecting decreased cortical slowing and enhanced neural activation. Relative power analysis indicated reduced Theta activity and Alpha normalization, suggesting improved cortical stability. Increases were observed in Beta and High-beta activity, alongside significant reductions in the Theta/Beta ratio, consistent with improved attentional regulation. Neuropsychological outcomes revealed reliable improvements in global cognition, memory, and visuospatial abilities, mostly maintained or enhanced at follow-up. Depressive and anxiety symptoms decreased markedly. Motor and functional assessments demonstrated meaningful improvements in motor performance, coordination, and functional independence. Conclusions: Findings suggest that integrating NFB with motor therapy may support recovery processes and be associated with sustained neuroplastic changes in the early post-injury phase after TBI, a condition associated with elevated risk for post-traumatic epilepsy. Full article

Obesity-associated inflammation underlies much of cardiometabolic pathology, reflecting the convergence of chronic, low-grade systemic immune activation with region-specific maladaptation of adipose depots. Among these, epicardial adipose tissue (EAT)—a visceral fat layer contiguous with the myocardium and sharing its microvasculature—functions as a cardio-proximal immunometabolic interface that influences atrial fibrillation, heart failure with preserved ejection fraction, and coronary atherogenesis through paracrine crosstalk. These relationships extend beyond crude measures of adiposity, emphasizing the primacy of local inflammatory signaling, adipokine flux, and fibro-inflammatory remodeling at the EAT–myocardium interface. Of importance, substantial weight reduction only partially reverses obesity-imprinted transcriptional and epigenetic programs across subcutaneous, visceral, and epicardial depots, supporting the concept of an enduring adipose memory that sustains cardiovascular (CV) risk despite metabolic improvement. Accordingly, therapeutic strategies should move beyond weight-centric management toward mechanism-guided interventions. Resolution pharmacology—leveraging specialized pro-resolving mediators and their cognate G-protein-coupled receptors—offers a biologically plausible means to terminate inflammation and reprogram immune–stromal interactions within adipose and CV tissues. Although preclinical studies report favorable effects on vascular remodeling, myocardial injury, and arrhythmic vulnerability, clinical translation is constrained by pharmacokinetic liabilities of native mediators and by incomplete validation of biomarkers for target engagement. This review integrates mechanistic, depot-resolved, and therapeutic evidence to inform the design of next-generation anti-inflammatory strategies for obesity-related CV disease. Full article

Dynamic edge-server allocation in mobile edge computing (MEC) networks is a challenging multi-objective optimization problem due to highly dynamic user demands, spatiotemporal traffic variations, and the need to simultaneously minimize service latency and workload imbalance. Existing heuristic and metaheuristic-based approaches for this problem often suffer from premature convergence, limited exploration–exploitation balance, and inadequate adaptability to dynamic network conditions, leading to suboptimal edge-server placement and inefficient resource utilization. Moreover, most existing methods lack memory-aware search mechanisms, which restrict their ability to capture long-term system dynamics. To address these limitations, this paper proposes a Fractional-Order Multi-Objective African Vulture Optimization Algorithm (FO-MO-AVOA) for dynamic edge-server allocation. By integrating fractional-order calculus into the standard multi-objective AVOA framework, the proposed method introduces long-memory effects that enhance convergence stability, search diversity, and adaptability to time-varying workloads. The performance of FO-MO-AVOA is evaluated using realistic MEC network scenarios and benchmarked against several well-established metaheuristic algorithms. Simulation outcomes reveal that FO-MO-AVOA achieves 40–46% lower latency, 38–45% reduction in workload imbalance, and up to 28–35% reduction in maximum workload compared to competing methods. Extensive experiments conducted on real-world telecom network data demonstrate that FO-MO-AVOA consistently outperforms state-of-the-art multi-objective optimization algorithms in terms of convergence behaviour, Pareto-front quality, and overall system performance. Full article

In a mother–child underwater bio-inspired robotic system, the equivalent lever arm between the master and slave inertial navigation systems (INSs) varies with launcher attitude changes and structural flexure. This time-varying lever arm introduces hard-to-model systematic velocity errors that degrade the accuracy and filter convergence of velocity difference-based transfer alignment. Traditional rigid body compensation relies on precise, constant lever-arm parameters and fails when booms, launch tubes, or flexible manipulators undergo appreciable deformation or reconfiguration. To address this, we augment a “velocity–attitude joint matching and innovation-based adaptive Kalman filter (AKF)” framework with an attention-based Long Short-Term Memory (LSTM) feed-forward module. Using only a short, real-time Inertial Measurement Unit (IMU) sequence from the slave INS, the module predicts and compensates the velocity bias induced by the lever arm. Numerical simulations of an underwater bio-inspired robot deployment scenario show that, under typical maneuvers (acceleration, turning, fin-flapping, and S-curve), the proposed method reduces the root-mean-square (RMS) misalignment angle error from about 14.5′ to 5.2′ and the RMS installation error angle from 8.8′ to 3.0′—average reductions of about 64% and 66%, respectively—substantially improving the robustness and practical applicability of transfer alignment under time-varying lever arms and flexible disturbances. Full article

Failures in clean water service lines can disrupt supply, increase operational costs, and reduce customer satisfaction. This study develops a machine learning framework to predict such failures, providing a proactive tool for utility asset management. A case study was conducted on a water distribution network in Tehran, serving approximately 205,000 customers, with 11 years of service line data and over 88,000 recorded failures. Service line attributes, including length, diameter, material, age, demand, and pressure, were combined with historical failure data to train Random Forest, Extreme Gradient Boosting, and Long Short-Term Memory models. Model performance was assessed using F1-score, AUC-ROC, and AUC-PRC. A novel metric was introduced to quantify failure reduction when prioritising replacements. The results demonstrate that machine learning can effectively capture complex relationships between service line features and failures, offering significant benefits for tactical maintenance planning. This research underscores the potential of predictive approaches to improve reliability and reduce costs. Full article