Search Results (7,418)

Digital twins are essential for the real-time health management and monitoring of aeroengines, and the in situ three-dimensional (3D) reconstruction technology of key components of aeroengines is an important support for the construction of a digital twin model. In this paper, an in situ high-fidelity 3D reconstruction method, named PoseNeRF, for aeroengine blades based on the joint optimization of pose and neural radiance field (NeRF), is proposed. An aeroengine blades background filtering network based on complex network theory (ComBFNet) is designed to filter out the useless background information contained in the two-dimensional (2D) images and improve the fidelity of the 3D reconstruction of blades, and the mean intersection over union (mIoU) of the network reaches 95.5%. The joint optimization loss function, including photometric loss, depth loss, and point cloud loss is proposed. The method solves the problems of excessive blurring and aliasing artifacts, caused by factors such as smooth blade surface and weak texture information in 3D reconstruction, as well as the cumulative error problem caused by camera pose pre-estimation. The PSNR, SSIM, and LPIPS of the 3D reconstruction model proposed in this paper reach 25.59, 0.719, and 0.239, respectively, which are superior to other general models. Full article

This study examines how cognitive biases may shape ethical decision-making in AI-mediated environments, particularly within education and research. As AI tools increasingly influence human judgment, biases such as normalization, complacency, rationalization, and authority bias can lead to ethical lapses, including academic misconduct, uncritical reliance on AI-generated content, and acceptance of misinformation. To explore these dynamics, we developed an LLM-generated synthetic behavior estimation framework that modeled six decision-making scenarios with probabilistic representations of key cognitive biases. The scenarios addressed issues ranging from loss of human agency to biased evaluations and homogenization of thought. Statistical summaries of the synthetic dataset indicated that 71% of agents engaged in unethical behavior influenced by biases like normalization and complacency, 78% relied on AI outputs without scrutiny due to automation and authority biases, and misinformation was accepted in 65% of cases, largely driven by projection and authority biases. These statistics are descriptive of this synthetic dataset only and are not intended as inferential claims about real-world populations. The findings nevertheless suggest the potential value of targeted interventions—such as AI literacy programs, systematic bias audits, and equitable access to AI tools—to promote responsible AI use. As a proof-of-concept, the framework offers controlled exploratory insights, but all reported outcomes reflect text-based pattern generation by an LLM rather than observed human behavior. Future research should validate and extend these findings with longitudinal and field data. Full article

Accurate inversion of the T₂ spectrum of shale oil reservoir fluids is crucial for reservoir evaluation. However, traditional nuclear magnetic resonance inversion methods face challenges in extracting features from multi-exponential decay signals. This study proposed an inversion method that combines autoencoder (AE) and Fourier transform, aiming to enhance the accuracy and stability of T₂ spectrum estimation for shale oil reservoirs. The autoencoder is employed to automatically extract deep features from the echo train, while the Fourier transform is used to enhance frequency domain features of multi-exponential decay information. Furthermore, this paper designs a customized weighted loss function based on a self-attention mechanism to focus the model’s learning capability on peak regions, thereby mitigating the negative impact of zero-value regions on model training. Experimental results demonstrate significant improvements in inversion accuracy, noise resistance, and computational efficiency compared to traditional inversion methods. This research provides an efficient and reliable new approach for precise evaluation of the T₂ spectrum in shale oil reservoirs. Full article

Artificial intelligence-based image processing is critical for sensor fusion and image transformation in mobility systems. Advanced driver assistance functions such as forward monitoring and digital side mirrors are essential for driving safety. Degradation due to raindrops, fog, and high-dynamic range (HDR) imbalance caused by lighting changes impairs visibility and reduces object recognition and distance estimation accuracy. This paper proposes a diffusion framework to enhance visibility under multi-degradation conditions. The denoising diffusion probabilistic model (DDPM) offers more stable training and high-resolution restoration than the generative adversarial networks. The DDPM relies on large-scale paired datasets, which are difficult to obtain in raindrop scenarios. This framework applies the Palette diffusion model, comprising data augmentation and raindrop-removal modules. The data augmentation module generates raindrop image masks and learns inpainting-based raindrop synthesis. Synthetic masks simulate raindrop patterns and HDR imbalance scenarios. The raindrop-removal module reconfigures the Palette architecture for image-to-image translation, incorporating the augmented synthetic dataset for raindrop removal learning. Loss functions and normalization strategies improve restoration stability and removal performance. During inference, the framework operates with a single conditional input, and an efficient sampling strategy is introduced to significantly accelerate the process. In post-processing, tone adjustment and chroma compensation enhance visual consistency. The proposed method preserves fine structural details and outperforms existing approaches in visual quality, improving the robustness of vision systems under adverse conditions. Full article

The paper presents the main features of a Web-GIS platform designed to compute real-time scenario-based seismic risk assessments at the national level. Based on the Italian experience, the platform enables DRM scientist and policymakers to readily generate seismic scenarios supporting the entire DRM cycle, including training, emergency planning, calibrating operations during response, and providing seismic risk estimates for National Disaster Risk Assessment or seismic risk reduction programs. The platform is immediately operational, relying on preloaded freeware datasets on exposure and vulnerability, and requiring only basic earthquake parameters to perform real-time analysis. At a later stage, these datasets should be replaced with more detailed and accurate national-level data. The platform generates earthquake impact assessments that include physical damage, economic and human losses, and key emergency response indicators, such as estimated displaced population, required tent camps, and EMT and USAR needs. Its key innovation lies in the ability to operate at the national scale, offering immediate usability with the possibility of further customization. As a web-based service with a user-friendly graphical interface, it is particularly suited for civil protection and DRM experts. Full article

Understanding how electromagnetic (EM) waves travel through different tissues is important for EM medical imaging, sensing, and in-body communication. It is known that EM waves in lossy bio-tissues are nonuniform and do not strictly follow the least time or least loss paths. Instead, they exhibit two distinct wavefronts: the phase wavefront and the amplitude wavefront, which are generally oriented at different angles. The impact of that on imaging and in-body communications is investigated and validated through comprehensive analysis and full-wave EM simulations. Additionally, the impact of a matching medium, commonly used to reduce antenna–skin interface reflections in medical EM applications, on the direction of EM wavefronts, travel time, phase changes, and attenuation is analyzed and quantified. The results show that the Fermat principle of least travel time, often used to estimate EM wave travel time for localization in medical imaging and wireless endoscopy, is only accurate when the loss tangent or dissipation factor of both the matching medium and tissues is very low. Otherwise, the results will be inaccurate, and the dual wavefronts should be considered. The presented analysis and results provide guidance on EM wave travel time and the direction of phase and amplitude wavefronts. This information is valuable for developing reliable processing algorithms for sensing, imaging, and in-body communication. Full article

Non-communicable diseases (NCDs) are the leading global cause of death, causing over 43 million deaths in 2021, including 18 million premature deaths, disproportionately affecting low- and middle-income countries. NCDs also incur significant economic losses, estimated at USD 7 trillion from 2011 to 2025, despite low prevention costs. This study evaluated body mapping indicators: body mass index (BMI), waist circumference, and waist-to-hip ratio—for predicting NCD risk, including hypertension, diabetes, and cardiovascular diseases, using data from a nationally representative survey in Ghana. The study sampled 5775 participants via multistage stratified sampling, ensuring proportional representation by region, urban/rural residency, age, and gender. Ethical approval and informed consent were obtained. Anthropometric and biochemical data, including height, weight, waist and hip circumferences, blood pressure, fasting glucose, and lipid profiles, were collected using standardized protocols. Data analysis was conducted with STATA 17.0, accounting for complex survey design. Significant sex-based differences were observed: men were taller and lighter, while women had higher BMI and waist/hip circumferences. NCD prevalence increased with age, peaking at 60–69 years, and was higher in females. Lower education and marital status (widowed, divorced, separated) correlated with higher NCD prevalence. Obesity and high waist circumference strongly predicted NCD risk, but individual anthropometric measures lacked screening accuracy. Integrated screening and tailored interventions are recommended for improved NCD detection and management in resource-limited settings. Full article

University campuses in rapidly urbanizing regions face increasing pressure to balance infrastructure development with environmental sustainability, yet their carbon storage potential remains largely unexplored in sub-Saharan Africa. This study assessed land use changes, carbon storage dynamics, and economic valuation across three Ghanaian universities, University of Ghana (UG), Kwame Nkrumah University of Science and Technology (KNUST), and University of Cape Coast (UCC), from 2017 to 2023, and evaluated five future scenarios using the InVEST carbon model. Land use analysis employed ESRI 10 m annual land cover data, while carbon storage was estimated using regionally appropriate carbon pool values, and economic valuation applied Ghana’s social cost of carbon ($0.970/tCO₂). Historical analysis revealed substantial carbon losses: UG declined by 17.1% (19,695 Mg C), KNUST by 29.5% (20,063 Mg C), and UCC by 7.9% (3292 Mg C), due to tree cover conversion to built areas. Scenario modeling demonstrated that infrastructure-focused development would cause additional losses of 4211–6891 Mg C, while extensive tree expansion could increase storage by 1686–5227 Mg C. Economic analysis showed tree expansion generating positive net present values ($1612–$5070), while infrastructure development imposed costs (−$4028 to −$6684). These findings provide quantitative evidence for sustainable campus planning prioritizing carbon conservation in tropical institutional landscapes. Full article

Cyberattacks on pipeline operational technology systems pose growing risks to energy infrastructure. This study develops a physics-informed simulation and optimization framework for analyzing cyber–physical threats in petroleum pipeline networks. The model integrates networked hydraulic dynamics, SCADA-based state estimation, model predictive control (MPC), and a bilevel formulation for stealthy false-data injection (FDI) attacks. Pipeline flow and pressure dynamics are modeled on a directed graph using nodal pressure evolution and edge-based Weymouth-type relations, including control-aware equipment such as valves and compressors. An extended Kalman filter estimates the full network state from partial SCADA telemetry. The controller computes pressure-safe control inputs via MPC under actuator constraints and forecasted demands. Adversarial manipulation is formalized as a bilevel optimization problem where an attacker perturbs sensor data to degrade throughput while remaining undetected by bad-data detectors. This attack–control interaction is solved via Karush–Kuhn–Tucker (KKT) reformulation, which results in a tractable mixed-integer quadratic program. Test gas pipeline case studies demonstrate the covert reduction in service delivery under attack. Results show that undetectable attacks can cause sustained throughput loss with minimal instantaneous deviation. This reveals the need for integrated detection and control strategies in cyber–physical infrastructure. Full article

The core challenge of the Unique Continuity (UC) problem lies in inferring solutions across an entire domain using limited observational data, holding significant practical implications for multiphysics coupled models. Recently, physics-informed neural networks (PINNs) have shown considerable promise in addressing the UC problem. However, the reliance on a fixed activation function and a fixed weighted loss function prevents PINNs from adequately representing the multiphysics characteristics embedded in coupled models. To overcome these limitations, we propose a novel dual adaptive neural network (DANN) algorithm. This approach integrates trainable adaptive activation functions and an adaptively weighted loss scheme, enabling the network to dynamically balance the observational data and governing physics. Our method is applicable not only to the UC problem but also to general forward problems governed by partial differential equations. Furthermore, we provide a theoretical foundation for the algorithm by deriving a generalization error estimate, discussing the potential causes of neural networks solving this problem. Extensive numerical experiments including 3D demonstrate the superior accuracy and effectiveness of the proposed DANN framework in solving the UC problem compared to standard PINNs. Full article

Hyperspectral imaging (HSI) systems often suffer from complex noise degradation during the imaging process, significantly impacting downstream applications. Deep learning-based methods, though effective, rely on impractical paired training data, while traditional model-based methods require manually tuned hyperparameters and lack generalization. To address these issues, we propose SS³L (Self-Supervised Spectral-Spatial Subspace Learning), a novel HSI denoising framework that requires neither paired data nor manual tuning. Specifically, we introduce a self-supervised spectral–spatial paradigm that learns noisy features from noisy data, rather than paired training data, based on spatial geometric symmetry and spectral local consistency constraints. To avoid manual hyperparameter tuning, we propose an adaptive rank subspace representation and a loss function designed based on the collaborative integration of spectral and spatial losses via noise-aware spectral-spatial weighting, guided by the estimated noise intensity. These components jointly enable a dynamic trade-off between detail preservation and noise reduction under varying noise levels. The proposed SS³L embeds noise-adaptive subspace representations into the dynamic spectral–spatial hybrid loss-constrained network, enabling cross-sensor denoising through prior-informed self-supervision. Experimental results demonstrate that SS³L effectively removes noise while preserving both structural fidelity and spectral accuracy under diverse noise conditions. Full article

This study investigates Italian fishery discards through the lens of neglected and underutilized species (NUS). It estimates the potential loss of value (PLoV) to identify pathways for sustainable valorization under the European Union landing obligation (LO). NUS were selected through a stakeholder focus group. Data regarding landings and discards were collected for the period 2020–2022 within the Italian Ministry of Agriculture, Food Sovereignty, and Forestry (MASAF) database. Among the three years, fleets landed roughly 130,400 tons annually, worth about €700 million, while discarding around 6200 tons yearly. This corresponds to an average PLoV of approximately €21.5 million. Most of the discarded quantity and value is concentrated in a few species. Atlantic Horse Mackerel stands out, accounting for nearly one-third of discarded biomass and about one-quarter of total PLoV. In 2020 and 2022, its discards even exceeded reported landings. A conservative valorization scenario for this single species indicates potential revenues of up to €7.5 million per year. Overall, these findings suggest that targeted NUS valorization could represent a way to diversify seafood consumption, alleviate pressure on common stocks, and buffer fishers’ incomes. This potential depends on ensuring traceability and safety, supported by pilots in processing, product development, and consumer acceptance. Full article

As a global staple ensuring food security, maize incurs 15–20% annual yield loss from pests/diseases. Conventional manual detection is inefficient (>7.5 h/ha) and subjective, while existing YOLO models suffer from >8% missed detections of small targets (e.g., corn armyworm larva) in complex fields due to feature loss and poor multi-scale fusion. We propose YOLO-FDLU, a YOLO11s-based framework: LAD (Light Attention-Downsampling)-Conv preserves small-target features; C3k2_DDC (DilatedReparam–DilatedReparam–Conv) enhances cross-scale fusion; Detect_FCFQ (Feature-Corner Fusion and Quality Estimation) optimizes bounding box localization; UIoU (Unified-IoU) loss reduces high-IoU regression bias. Evaluated on a 25,419-sample dataset (6 categories, 3 public sources + 1200 compliant web images), it achieves 91.12% Precision, 92.70% mAP@0.5, 78.5% mAP@0.5–0.95, and 20.2 GFLOPs/15.3 MB. It outperforms YOLOv5-s to YOLO12-s, supporting precision maize pest/disease monitoring. Full article

The first comprehensive mass balance estimation for the world’s largest tropical icefield is presented. Geodetical mass balance was calculated using photogrammetry from aerial and satellite images spanning from 1955 to 2024. The results meet expected quality standards using some new satellite sources, such as the Peruvian PeruSAT-1, although the quality of airborne imagery is consistently lower than that of satellite sources. The Nevado Coropuna icefield remained almost stable between 1955 and 1986 with −0.04 m dh yr⁻¹. Since then, it has undergone a sustained and accelerated negative mass balance, reaching a maximum annual dh yr⁻¹ of −0.73 ± 0.19 in the 2020–2023 timeframe. The glacier loss is not equal across the entire ice mass, but more acute in the northern and northeastern outlet tongues. Debris-covered ice and rock glaciers show a much weaker negative mass balance signal. The impact of ENSO events is not evident in the overall ice evolution, although their long-term relevance is acknowledged. Overall, the negative response of Nevado Coropuna to global warming (−0.36 ± 0.12 m.w.e. yr⁻¹ for the 2013 to 2024 period) is less pronounced than that of other Peruvian glaciers, but more severe than that reported for the nearby Dry Andes of Chile and Argentina. Full article

Power converters are increasingly pushing toward higher switching frequencies, with current designs typically operating between tens of kilohertz and a few megahertz. The commercialization of gallium nitride (GaN) power transistors has opened new possibilities, offering performance far beyond the limitations of conventional silicon devices. Despite this promise, the potential of GaN technology remains underutilized. This paper explores the feasibility of achieving sub-gigahertz switching frequencies using GaN-based switch-mode power converters, a regime currently inaccessible to silicon-based counterparts. To reach such operating speeds, it is essential to understand and quantify the intrinsic frequency limitations imposed by GaN device physics and associated parasitics. Existing power conversion topologies and control techniques are unsuitable at these frequencies due to excessive switching losses and inadequate drive capability. This work presents a detailed, systematic study of GaN transistor behavior at high frequencies, aiming to identify both fundamental and practical switching limits. A compact analytical model is developed to estimate the maximum soft-switching frequency, considering only intrinsic device parameters. Under idealized converter conditions, this upper bound is derived as a function of internal losses and the system’s target efficiency. From this, a soft-switching figure of merit is proposed to guide the design and layout of GaN field-effect transistors for highly integrated power systems. The key contribution of this study lies in its analytical insight into the performance boundaries of GaN transistors, highlighting the roles of parasitic elements and loss mechanisms. These findings provide a foundation for developing next-generation, high-frequency, chip-scale power converters. Full article