Search Results (3,671)

Eye tracking is a valuable behavioral signal for human-centered AI, yet the reliability of feature-based machine-learning models for inferring task type and workload across users and tasks remains uncertain, because experimentally defined workload labels may reflect task type and visual structure as much as cognitive demand. The practical problem is that designers of gaze-adaptive systems need to know which inferences are dependable enough to act on, and reported accuracies alone do not answer this, because the choice of prediction target and validation split can determine the result. This study systematically evaluates feature-based machine-learning models with post hoc explainability across three prediction targets: task type, binary load-versus-rest, and three-level workload. Eye-movement features derived from fixations, saccades, pupils, and blinks were extracted from short temporal windows collected from 54 participants performing attention, visual-spatial, and memory tasks under rest, easy, and difficult conditions, and evaluated using leave-one-subject-out (LOSO) and leave-one-group-out (LOGO) validation. Task type was classified most reliably (85.9% LOSO, 83.4% LOGO), binary load-versus-rest showed moderate, validation-sensitive robustness (81.4% LOSO, 63.9% LOGO), and three-level workload classification was substantially more challenging (56.4% LOSO, 44.3% LOGO). SHAP and statistical analyses consistently identified fixation dispersion, pupil-related measures, and subject-normalized features as the strongest contributors across all three targets. These findings show that prediction target definition, validation strategy, and post hoc explainability jointly determine what can be reliably inferred from gaze-based machine-learning models. Eye tracking alone therefore appears promising for task-type recognition and may support coarse engagement-related inference when the deployment task family is represented during model development, whereas task-independent fine-grained workload estimation remains unsupported by the present evidence. Full article

High-bandwidth memory (HBM) faces critical power integrity challenges in high-stack configurations due to elongated power delivery paths and increased parasitic inductance. This paper proposes a dual-interposer HBM architecture with an interposer–HBM stack–interposer configuration, integrating an additional top interposer embedded with chip capacitors. This topology redesigns the HBM’s power distribution network, reducing PDN impedance, and this technology enables bidirectional vertical power supply to DRAM chips during moments when they require current. The PDN impedance is systematically compared with a conventional trench-capacitance-enhanced structure (Structure A) and a deep-trench-capacitance-enhanced structure (Structure B). Results show that at 0.1–11.2 GHz, the proposed structure reduces peak PDN impedance by 66.41% and 65.7% versus Structures A and B, respectively, and decreases the loop inductance of the top-layer DRAM chip by 66.71%. The top interposer’s redistribution layer forms a parallel-plate capacitor complementing the embedded chip capacitors, achieving wideband impedance suppression. Without modifying existing protocols, this architecture provides a system-level PDN optimization strategy for high-stack HBM, offering quantitative insights for capacitor selection and layout design. Full article

Endovascular aortic aneurysm repair (EVAR) has become the first-line treatment for abdominal aortic aneurysms (AAA); however, its long-term durability is affected by Type II endoleak. This narrative review describes the evolution of sac filling techniques and concepts developed to address this complication. In selected cases of diagnosed Type II endoleaks, early therapeutic sac filling, while clinically indicated, is frequently constrained by difficult access, variable success rates, and recurrence. Consequently, prophylactic sac filling during primary EVAR has been adopted as an alternative approach, with reported high technical success and reduced endoleak incidence. The subsequent Nellix Endovascular Aneurysm Sealing (EVAS) system, although early results showed technical feasibility, was associated with graft migration, endobag separation, and reintervention at long-term follow-up, leading to its withdrawal from the market. Currently, devices based on shape memory polymer (SMP) provide early observations of decreased endoleaks and sac regression. This review outlines the transition from reactive management to preventive strategies, and, more recently, to patient-specific, material-based approaches, while showcasing emerging strategies that may influence long-term EVAR outcomes. However, these novel approaches remain under investigation, and larger controlled studies with extended follow-up are required to establish their clinical utility. Full article

Meningitis is a significant public health problem despite the availability of effective vaccination programs, especially in children and young people. The memory-dependent features of disease transmission, immunity and vaccination dynamics are not often represented in classical integer-order epidemic models. This study proposes a fractional-order model of meningitis transmission with memory using the Atangana–Baleanu–Caputo fractional derivative. The model features susceptible, vaccinated, exposed, infectious, treated, and recovered populations to assess the impact of vaccination coverage, vaccine effectiveness, loss of vaccine immunity, and treatment on the spread of meningitis. The basic mathematical characteristics of the model, such as positivity, existence, uniqueness, and stability of solution are proven. The Laplace–Adomian Decomposition Method (LADM) is used to obtain the approximate analytical solutions, and a numerical simulation is used to analyze the influence of the fractional-order memory and epidemiological parameters on the epidemic process. The most important parameters that influence the basic reproduction number are found in sensitivity analysis to be the transmission rate and the vaccination-related parameters. The results show that simply increasing the vaccination coverage and vaccine effectiveness can substantially decrease the number of disease transmissions, and vaccine coverage can produce memory effects to change the timing and duration of outbreaks. The suggested fractional-order computational framework is a framework that is vital for studying the dynamics of meningitis and can be used for the design of long-term vaccination and disease-control strategies. Full article

Solar tracking can increase photovoltaic (PV) energy yield, but its contribution to sustainable electricity depends on more than geometric gain. This structured narrative review evaluates flat-plate and low-concentration PV trackers using an integrated three-layer taxonomy covering mechanical architecture, actuation and drivetrain, and control strategy. Tracker classes are compared in terms of annual energy gain, life-cycle cost, parasitic consumption, land-use efficiency, structural resilience, reliability, maintainability, and deployment maturity. Utility-scale horizontal single-axis trackers using astronomical control, backtracking, supervisory monitoring, and weather-dependent stow provide the most mature balance of energy yield, cost, and operational robustness. Dual-axis systems can offer higher output under high-direct-normal-irradiance conditions but impose greater structural and O&M burdens, while passive fluid-based and shape-memory-alloy concepts remain mainly experimental. The review also examines bifacial and terrain-aware tracking, agrivoltaic dual land use, extreme-weather resilience, tracker-specific availability, artificial intelligence, digital twins, predictive maintenance, and end-of-life considerations. A plant-level decision framework and reporting checklist are proposed to support transparent, project-specific choices that maximize lifetime renewable-energy value while limiting material use, land-use conflict, operational risk, and avoidable life-cycle impacts. Full article

Pre-trained language models are typically adapted to downstream tasks via full fine tuning. However, this entails substantial computational and memory overhead. Parameter-efficient alternatives such as Low-Rank Adaptation (LoRA), Quantized Low-Rank Adaptation (QLoRA), and frozen backbone mitigate these costs, but empirical evidence on how their performance–cost trade-offs change under low labeled data in Arabic remains limited. This paper compares four adaptation strategies: full fine tuning, frozen backbone, LoRA, and QLoRA for Arabic binary sentiment classification on the Hotel Arabic Reviews Dataset, using CAMeLBERT-Mix as the pre-trained encoder. The methods are evaluated under a unified experimental setting at three labeled-data levels: the full training set, 100 samples per class, and 25 samples per class. The evaluation metrics are reported as means and standard deviations across five random seeds. At the full-data level, full fine tuning, LoRA, and QLoRA achieve macro-F1 scores between 0.9569 and 0.9579 and are comparable within seed variability, while the frozen backbone exhibits performance that is approximately ten points lower. LoRA and QLoRA use approximately 35.0% less peak GPU memory than full fine tuning but require longer training times. Under reduced-data conditions, full fine tuning outperforms all other adaptation strategies with the differences being statically significant. Full article

This paper presents a novel fractional-order hybrid observer framework for robust sensorless control of brushless DC (BLDC) motor drives in unmanned aerial vehicle (UAV) propulsion systems, addressing the fundamental limitations of conventional integer-order observers through the lens of fractional calculus. The proposed architecture synergistically integrates high-frequency square-wave signal injection for zero/low-speed operation with an adaptive fractional-order extended Kalman filter (AFEKF) augmented by online stator resistance and flux linkage estimation, capitalizing on the memory and hereditary properties inherent to fractional-order systems. A minimum-order current observer enables accurate three-phase current reconstruction using a single DC-link sensor, substantially reducing hardware complexity and cost. The complete algorithm is implemented on an STM32H7 microcontroller and experimentally validated on a 1.5 kW drone propulsion testbench and in-flight platform. Results demonstrate reliable startup under 50% rated load, stable operation from standstill to 5000 RPM on the UAV motor (and validated up to 22,000 RPM on a high-speed test motor, <4° electrical position error at 5 kRPM, and strong robustness against 35% stator resistance variation. In-flight tests confirm improved thrust smoothness and hover stability compared to conventional sensorless strategies. The proposed fractional-order architecture offers a practical, resilient, and computationally feasible solution for next-generation autonomous aerial systems, establishing a new paradigm for observer design in electric propulsion. Full article

The dissimilar combination of NiTi shape memory alloy and Ti6Al4V titanium alloy offers superelasticity, biocompatibility and high specific strength, showing broad application prospects in aerospace and medical fields. However, when fabricating NiTi/Ti6Al4V composite components by multi-material laser powder bed fusion (MM-LPBF), brittle cracks or even complete delamination easily occur at the interface. In this paper, without relying on intermediate interlayer materials, we innovatively propose a root-inspired three-dimensional bio-interlocking interface structure. By means of macroscopic three-dimensional geometric interlocking, the crack propagation path and load transfer mode are forced to change. Using the branching angle (45°, 60°) and the structural size multiplier (1.2, 1.5) as variables, the influence of the bio-inspired geometric parameters on the interfacial forming quality, microstructure and mechanical properties was systematically investigated. The results show that the branching angle is the primary factor determining the performance. The 45° low-angle branched specimens exhibit overall brittle delamination along the flat metallurgical reaction interface under shear loading, with an average shear strength of only 17.47 MPa. In contrast, the 60° high-angle branched specimens, owing to their larger normal embedding depth, exhibit a failure mode transitioning to a mixed mode that includes crack deflection, branch shearing and plastic tearing of the Ti6Al4V matrix. Although TEM confirms that a continuous Ti₂Ni brittle phase still exists at the interface, the optimised 60–1.5 structure increases the average shear strength to 128.37 MPa, which is more than six times higher than that of the 45–1.2 group (17.47 MPa). This “geometrical constraint toughening” strategy provides a new paradigm for the interfacial strengthening of dissimilar metals without relying on metallurgical modification. Full article

Multi-modal 3D object detection is an important task in autonomous driving systems, where cameras and LiDAR provide complementary semantic and geometric information. Most existing BEV fusion methods are designed based on the Cartesian representation space, which does not fully match the sensing geometry of camera and LiDAR. This generally leads to redundant computation in distant regions. To address this issue, GARF, a geometry-aware polar BEV framework, is presented for multi-modal 3D object detection. GARF organizes camera and LiDAR features in a unified polar BEV space, which can represent spatial resolution more compactly. For the camera branch, the uncertainty-guided transformation of the polar view is designed to improve the reliability of depth estimation. Then, the generated polar BEV feature is further refined to attenuate radial noise and angular discontinuity. For the LiDAR branch, the polar-aware sparse feature extraction and distortion correction modules are designed to deal with the anisotropic structure and geometric distortion caused by polar voxelization. For multi-modal fusion, the region-aware cross-modal fusion strategy and polar detection head with anisotropic Gaussian center response map are developed, which achieve effective feature interaction and consistent geometry supervision. The experimental results on nuScenes show that GARF achieves 71.8% mAP and 73.7% NDS, improving the baseline by 3.3% mAP and 2.3% NDS. Meanwhile, the inference speed increases from 7.1 FPS to 8.9 FPS, and the consumption of GPU memory decreases from 41,114 MiB to 33,346 MiB. Full article

Magnetic shape memory alloys based on the Ni–Mn–Ga system are of strategic interest for aerospace and robotics applications due to their ability to respond to both thermal and magnetic stimuli. However, the NASA Shape Memory Materials Database a key resource for the community exhibits significant gaps in functional parameters, with up to 93.7% of records missing critical properties such as the Curie temperature, and over 88% lacking complete magnetic data. To address this limitation, this study proposes a data imputation strategy based on a stacking ensemble comprising twelve machine learning models (LGBM, XGBoost, CatBoost, GradientBoosting, RandomForest, MLP, BayesianRidge, KNN, SVR, GPR, MICE, and AutoEncoder), optimized via Optuna and evaluated using ten random seeds with 10 repetitions each. The approach was applied to reconstruct missing entries in NASA’s database. For heat treatment 1, the method achieved coefficients of determination ($R^2$) of 0.95 for duration (h) and 0.88 for temperature (°C), respectively. For the phase transformation temperatures (Mf, Ms, As, and Af), the method yielded $R^2$ values of 0.83, 0.82, 0.79, and 0.80, respectively. Magnetic properties saturation magnetization and maximum magnetic field were imputed with an $R^2$ of 0.92. In contrast, the Curie temperature exhibited limited predictive performance ($R^2$ = 0.15–0.35), primarily due to insufficient data availability. Overall, the proposed methodology integrates machine learning based imputation with physically supported constraints, providing a viable alternative to enhance the completeness and utility of materials databases. Full article

The exponential growth of the Internet of Things (IoT) has generated massive data streams traditionally processed by centralized cloud architectures, which increasingly face latency, bandwidth, and privacy limitations. Shifting artificial intelligence to resource-constrained edge nodes, known as TinyML, offers a robust decentralized alternative, though it introduces severe memory, compute, and energy bottlenecks. To map this transition, a systematic literature review was conducted following PRISMA guidelines, analyzing peer-reviewed studies published between 2021 and 2026 across major databases. The analysis identifies primary architectural paradigms and evaluates the efficacy of state-of-the-art model compression techniques, such as quantization, pruning, and knowledge distillation. Furthermore, the findings reveal that hardware–software co-design and custom neural accelerators are crucial for overcoming operational bottlenecks, while also highlighting persistent security and privacy challenges in on-device learning. Ultimately, while deploying complex models on microcontrollers is increasingly viable, achieving optimal performance demands holistic optimization strategies. This review synthesizes current research gaps and provides a strategic roadmap to guide future interdisciplinary efforts toward resilient, energy-efficient, and secure next-generation intelligent edge systems. Full article

Accelerometer-based tilt angle estimation is widely used in engineering monitoring, yet random noise and outliers degrade its accuracy. Integer-order Kalman filters suppress noise, but their Markovian model assumes that the current state alone is sufficient to predict the next, neglecting the influence of earlier states on slowly varying processes. Fractional-order Kalman filters incorporate historical states into the prediction. However, the conventional formulation shares a single transition matrix between the state and covariance predictions, underestimating the prediction uncertainty, while the memory mechanism propagates gross errors across iterations. To overcome these limitations, this paper proposes a decoupled fractional-order Kalman filter (DFKF). The method assigns independent transition matrices to the state and covariance predictions, where a scaling coefficient inflates the predicted covariance to lower the prediction weight and strengthen reliance on measurements. A front-end gross-error pre-elimination strategy combining second-order differencing with adaptive peak detection is further introduced to block outlier propagation at the source before it enters the memory mechanism. Simulations under varying noise levels and gross-error conditions show that DFKF achieves a mean RMSE (Root Mean Square Error) of 0.147°, representing reductions of 17.4% and 6.4% over KF (0.178°) and FKF (0.157°), respectively, and a mean MaxAE (Maximum Absolute Error) of 0.548°, outperforming KF and FKF by 24.5% and 10.3%. Under gross-error conditions, DFKF converges in 0.012 s on average, approximately 2.8 times faster than KF and FKF, and the pre-elimination strategy restores accuracy to near-error-free levels. Full article

Reliable traffic flow forecasting is a core component of intelligent transportation systems; however, many current approaches are still unable to simultaneously model spatial interdependencies and long-term temporal correlations, particularly in cross-sea corridors that exhibit directional heterogeneity and pronounced temporal variability. This study aims to develop an accurate and stable traffic flow prediction framework for cross-sea corridors. To achieve this, an HGTransformer model was proposed that constructed a hypergraph from traffic nodes based on spatial proximity and correlated flow variations, and used hypergraph convolution to extract spatial node representations. These representations were then fed into a Transformer equipped with multi-head self-attention and positional encoding, enabling the model to capture global temporal dependencies in the evolution of traffic flow. Using hourly traffic flow data from the Huangmaohai cross-sea corridor, the model was tested on 1 to 4 h forecasting horizons and compared with long short-term memory (LSTM), multi-layer perceptron (MLP), random forest (RF), support vector regression (SVR), and Bayesian regression (BR) models. The proposed model achieved the best overall performance, with average mean absolute percentage error (MAPE), mean absolute error (MAE), weighted mean absolute percentage error (WMAPE), and root mean square error (RMSE) of 0.178, 13.375, 0.140, and 20.538, respectively. At the 1 h horizon, these values further decreased to 0.172, 12.678, 0.132, and 19.401, while preserving peak–valley structures more accurately under both short- and longer-horizon forecasting. The main contribution of this study lies in the systematic application and validation of the Huangmaohai Corridor dataset, including a reproducible hypergraph construction strategy tailored specifically for this particular infrastructure. Full article

This paper presents an adaptive noise-aware bearing fault diagnosis framework that integrates Fast Fourier Transform (FFT)-windowed feature extraction, wavelet-based Signal-to-Noise Ratio (SNR) estimation, and Long Short-Term Memory (LSTM) classification to maintain high diagnostic accuracy under both clean and severely noisy operating conditions. The core novelty of the proposed framework lies in its adaptive model selection mechanism, which automatically selects the most appropriate LSTM classifier according to the estimated SNR, thereby improving diagnostic robustness across different noise environments. Experiments were conducted on two benchmark datasets, the Case Western Reserve University (CWRU) dataset and the Huazhong University of Science and Technology (HUST) bearing dataset, to evaluate the generalization capability of the proposed approach. Two preprocessing pipelines were examined: time-domain normalization before FFT and frequency-domain normalization after FFT. Vibration signals were segmented without overlap to ensure unbiased evaluation. The results demonstrate that both the choice of window function and the normalization strategy significantly influence classification accuracy and robustness. Under noise-free conditions, several window types achieved accuracies above 99%, with triangular and Hamming windows providing the best performance. The combination of triangular windowing and time-domain normalization achieved the highest accuracy of 99.69%. Furthermore, time-domain normalization combined with triangular windowing exhibited superior stability and noise resistance compared with frequency-domain normalization. Under noisy conditions, noise-augmented training was found to be essential for achieving robust generalization. Models trained with moderate noise levels (8–12 dB) provided the best trade-off between accuracy and robustness, whereas excessive noise during training degraded performance. To accommodate varying noise environments, a lightweight wavelet-based SNR estimator was used to categorize operating conditions into low-, medium-, and high-SNR regions and select the corresponding LSTM classifier. The proposed framework was successfully implemented on a ZedBoard FPGA (Field-Programmable Gate Array) development board using a System-on-Chip (SoC) architecture. Experimental results show that, with a sampling frequency of 48 kHz and a processing window of 2048 samples, the proposed system updates the diagnostic result every 42.7 ms, demonstrating its suitability for real-time industrial condition monitoring and intelligent predictive maintenance applications. Full article

SPARQ is an advanced global optimization algorithm, the direct evolution of an earlier method called ARQ2. It searches efficiently for the best solution to problems where the space of options is too vast to check exhaustively, such as tuning antenna arrays or planning fuel-efficient spacecraft trajectories. Like its predecessor, it works with a population of candidate solutions that improve generation after generation, alternating between two complementary search strategies. What sets SPARQ apart is that it makes nearly every part of this process adaptive. Its population shrinks intelligently as the search matures. Its internal settings draw from a memory of many past successful configurations, not a single average. Its escape-from-stagnation mechanisms come in graduated strength, from a gentle nudge to a deeper partial restart. It also adds capabilities its predecessor never had, including a dedicated phase that locally polishes the current best solution using its own memory of productive directions. Every addition is kept only where it showed an overall benefit during development, though a subsequent component-wise analysis shows this benefit varies markedly in size and statistical significance across mechanisms. The result is more reliable than its predecessor on the large majority of tested problems, while staying grounded in the same battle-tested core. Full article