Search Results (7,801)

Accurate characterization of riverbed substrate from remote sensing imagery is essential for applications in fluvial geomorphology, habitat modeling, and river management. While recent advances in computer vision, particularly deep learning, have improved sediment mapping capabilities, their reliance on large annotated datasets and computational resources limits their broader applicability. This study presents a scalable workflow for categorical substrate classification using ultra-high-resolution aerial RGB orthoimagery in clear-water river environments. The approach integrates spectral information with statistical and structural texture descriptors derived from Gray-Level Co-occurrence Matrices (GLCM) and Local Binary Patterns (LBP), combined within a Random Forest classification framework. The methodology is structured as a semi-automated, five-stage workflow: (1) expert-based ground-truth substrate annotation; (2) feature set generation; (3) spatially aware model optimization; (4) full-domain classification; and (5) design-based validation for independent accuracy assessment. Model performance is evaluated using spatially aware cross-validation and design-based probability sampling to account for spatial autocorrelation and provide unbiased accuracy estimates. The method was applied in four geomorphologically distinct alpine river reaches, achieving design-based overall accuracy ranging from 70% to 88%. These results demonstrate that RGB-based approaches can achieve reliable reach-scale categorical substrate classification when combined with appropriate feature representation and rigorous validation strategies. However, limitations remain for visually similar or transitional substrate classes, particularly fine sediments such as sand and clay, which are difficult to distinguish consistently even during manual annotation. The workflow is implemented using open-source tools and is applicable to clear-water conditions where the riverbed remains optically visible. Full article

Post-event building damage assessment from bi-temporal satellite imagery is challenging because damaged pixels are sparse, structural changes can resemble radiometric variation, and sensor changes can dominate direct feature subtraction. This study evaluates SiameseCTDM with a Sensor-Adaptive Cross-Temporal Difference Module (SA-CTDM). At each encoder stage, SA-CTDM combines absolute and magnitude-normalized relative discrepancies through a learned per-channel gate, followed by channel recalibration. The gate provides adaptive interpolation between two discrepancy definitions; it is not treated as an independently superior accuracy component. A source-preserving class-conditional objective is also evaluated for optical-to-SAR adaptation by combining supervised BRIGHT learning, continued xBD supervision, and multi-stage alignment of intact and damaged prototypes. The adaptation terms are training-only and add no inference-time parameters. On the xBD hold set, complete SA-CTDM obtains 66.68 ± 0.54% F1, compared with 66.16 ± 1.12% for original CTDM, representing a modest 0.52 percentage-point mean increase with lower variance. A normalized-only ablation is the strongest single run, so an independent gating advantage is not claimed. A unified batch-one benchmark at $512\times 512$ input resolution reports measured latency and throughput in addition to parameter counts and GFLOPs. On a fixed seed-42 sample-level BRIGHT split, the complete configuration using 611 labeled pairs reaches 34.33% F1, whereas the aligned 305-pair configuration does not outperform vanilla fine-tuning. The results therefore characterize split-specific and label-budget-dependent behavior rather than uniform few-shot superiority. Full article

Three-dimensional (3D) graphene foams are attractive as lightweight conductive scaffolds with large surface area and broadband light absorption but achieving reproducible porosity and preserving the architecture after metal-template removal remain challenging. Here we report a stepwise route to freestanding 3D multilayer graphene foams based on (i) hydrogen-bubble-assisted electrodeposition of porous Ni on Cu foils, (ii) time-controlled pre-annealing at 1000 °C to drive Cu diffusion and form porous Ni-Cu alloy templates, (iii) in situ graphene CVD at 1000 °C under fixed growth conditions, and (iv) wet etching to remove the metal scaffold without a polymer support. The influence of pre-annealing (0, 1, 3, and 7 h) on template evolution, graphene growth, and foam stability was systematically investigated via SEM, EDS, XRD and Raman studies. Before etching, Raman spectroscopy indicates low-defect graphenic coatings with locally heterogeneous few-layer-like to multilayer-like signatures. Only samples pre-annealed for at least 3 h preserved the porous 3D architecture after metal removal, indicating the formation of self-supporting graphenic networks with improved post-etch morphological stability. Raman and XRD analyses further revealed a progressive reduction in structural degradation, residual strain, and stacking disorder with increasing pre-annealing time. Among the investigated samples, the foams obtained after 3 and 7 h of template pre-annealing combined preserved 3D morphology with low sheet resistance (10–20 Ω/□), negligible optical transmittance (<5%), and strong broadband visible-light absorption (75–90%). Full article

Biosensors have become important analytical platforms that enable rapid, selective, sensitive and portable analysis for early disease diagnosis, biomarker monitoring and point-of-care diagnostic applications. Their analytical performance depends on the coordinated function of molecular recognition elements, surface chemistry, transduction mechanisms and signal-processing strategies. Nevertheless, the analysis of real biological samples remains challenging because of low target concentrations, matrix effects, interfering species, signal noise, sensor drift and device-to-device variability. Therefore, artificial intelligence and machine learning are gaining increasing importance as data-driven tools for signal preprocessing, calibration, feature extraction, pattern recognition, quantitative prediction and diagnostic decision support. These approaches are particularly valuable for interpreting complex datasets generated by electrochemical, optical, wearable and microfluidic biosensors. This review presents an overview of healthcare-oriented biosensor systems beginning with molecular recognition principles, bioreceptor design, and transduction technologies, and extending to applications in clinical diagnosis and health monitoring. It also examines the roles of supervised, unsupervised and deep learning approaches in biosensor data analysis, while critically discussing model validation, generalizability, interpretability and clinical translation. By linking molecular-level recognition with computational signal interpretation, this review highlights the advantages and limitations of artificial intelligence-integrated biosensors for next-generation point-of-care diagnostics, continuous health monitoring, and personalized healthcare applications. Full article

The chiral interaction between light and matter is intrinsically weak at the nanoscale, which leads to a sensitivity bottleneck in the detection of chiral samples by photoinduced force microscopy. To address this issue, we propose an enhancement strategy that combines a semiconductor tip with a superchiral optical field to improve the optical force response in nanoscale chiral detection. Compared with a conventional metallic tip, a silicon tip can generate a larger optical force difference under left- and right-handed circularly polarized illumination. This enhancement mainly arises from the more pronounced magnetic response of the silicon tip, which increases the contribution of the magnetic dipole term to the total optical force. Furthermore, when the incident field is changed from a circularly polarized field to a superchiral field, the optical force difference acting on the same tip can be further increased by more than one order of magnitude. Relative to the reference case of a gold tip excited by circularly polarized light, the combination of a superchiral field and a silicon tip enhances the optical force difference by nearly two orders of magnitude. Analysis based on the dipole approximation shows that this enhancement originates from the synergistic effect of two factors: the high-refractive-index silicon tip provides a stronger magnetic response, while the superchiral field further strengthens the coupling between the localized chiral field and the tip and effectively excites the magnetic dipole and electromagnetic coupling terms, thereby jointly amplifying the chiral optical force signal. This work provides a new route for the highly sensitive detection of chiral materials at the nanoscale and may further extend the applications of photoinduced force microscopy in the characterization of chirality and optomagnetic interactions. Full article

High-entropy oxide (HEO) nanoceramics based on (Y_0.2La_0.2Gd_0.2Eu_0.2Er_0.2)₂O₃ were synthesized at 600 °C for 10 min under pressures of 2, 4, 6, and 8 GPa using High-Pressure–Low-Temperature (HPLT) technology, and the resulting structural, phase, and optical changes were studied as a function of synthesis pressure. X-ray diffraction with Rietveld refinement showed that the initial single-phase cubic nanopowder decomposes under pressure into a mixture of cubic, monoclinic, and orthorhombic high-entropy phases: the cubic fraction falls from 67.3% at 2 GPa to 46.8% at 4 GPa, 42.9% at 6 GPa, and 31.1% at 8 GPa, while low-symmetry monoclinic and orthorhombic inclusions become correspondingly more abundant. Raman spectroscopy validated this evolution, with the 2 GPa sample showing a resolvable doublet near 364 and 353 cm⁻¹ attributable to two coexisting cubic phases, while samples synthesized at 4–8 GPa converge on a single narrow band at 353 cm⁻¹ that broadens with increasing pressure. Photoluminescence measurements revealed that the sample synthesized at 2 GPa exhibits the highest Eu³⁺ and Er³⁺ luminescence intensity, with emission and excitation intensities decreasing systematically as synthesis pressure increases. We attribute this decline to the growing fraction of low-symmetry monoclinic phase, whose C2h point-group sites impose parity-forbidden selection rules on the ⁵D₀ → ⁷F_J transitions of Eu³⁺, combined with an increased probability of nonradiative relaxation at structural defects introduced by pressure. These results establish synthesis pressure as a practical lever for tuning the phase composition and luminescent efficiency of multi-lanthanide HEO nanoceramics and indicate that low pressures (~2 GPa) are preferable for optical applications requiring high luminescence intensity. Full article

The concentrations of 18 metal(loid)s were measured in five marine invertebrate species (Oxystele tigrine and sinensis, Marthasterias glacialis, Cymbula oculus and granatina, Parechinus angulosus, and Mytilus galloprovincialis) and five macroalgae/seaweed species (Aeodes orbitosa, Gelidium pristoides, Caulerpa filiformis, Ulva sp., and Bifurcaria brassicac formis) collected from eight sites within the marine environment of False Bay, Cape Town, South Africa. To assess the potential human health hazards associated with consumption, samples were acid digestion and analysed using inductively coupled plasma optical emission spectrometry (ICP-OES). The analysed elements included trace metals (Fe, Mn, Zn) toxic metals (Pb, Cd, As), transition metals (Ta, Ti), alkali metals (Li), and metalloid (Se, As). Metal concentration varied widely among species and locations, with values ranging from Zn: not detected (nd)—3836.04 mg/kg dry weight (dw); As: nd—77.98 mg/kg dw; Pb: nd—301.15 mg/kg dw; Li: nd—97.28 mg/kg dw, indicating diverse metal accumulation patterns. Human health risk was evaluated using the target hazard quotient (THQ) and hazard index (HI). Arsenic consistently exceeded the THQ threshold of one across all sites, and all the HI values were greater than one for all edible species, suggesting potential non-carcinogenic health risks associated with the consumption, except for M. glacialis, which is not commonly consumed. The HI metrics demonstrate the adverse impacts of metal concentration in marine biota. The observed spatial and biological variability in metal accumulation highlights the limitations of relying solely on metal concentrations for assessing health risks, food safety or anthropogenic contamination, and therefore more frequent and comprehensive monitoring is recommended. Full article

Multi-class oriented ship detection in optical remote sensing images remains challenging in densely berthed and nearshore scenes. Elongated hulls, arbitrary headings, background clutter, and similar vessel appearances can weaken feature aggregation and reduce the accuracy of rotated-box localization. This study proposes YOLO-CPCL, a compact oriented detector developed from YOLOv8n-OBB. In the final YOLO-CPCL architecture, a Ship-Oriented Slender Adaptive Fusion module (SOSA-Fuse) replaces all four C2f fusion units in the Neck. It combines learned content-adaptive sampling with a C2f-style split-and-concatenation pathway to improve feature aggregation for elongated and arbitrarily oriented ships. An Aspect-Ratio-Coupled Angle Supervision Loss (ARCAS-Loss) is further introduced by applying a bounded logarithmic aspect-ratio weight to a periodic cosine angle term. This formulation strengthens angle supervision for slender targets while limiting the influence of extreme samples. On the four-class HRSC2016 task, YOLO-CPCL improves precision, recall, mAP@50, mAP@75, and mAP@50–95 by 3.54, 7.39, 5.48, 8.60, and 7.23 percentage points, respectively. The parameter count is reduced from 3.08 M to 2.92 M, corresponding to a decrease of 5.19%, while the computational cost is reduced from 8.3 to 7.6 GFLOPs, a decrease of 8.43%. Additional evaluations on the Level-2 24-class setting of ShipRSImageNet and a custom DOTA-v1.0 protocol with multi-class training and ship-class reporting show positive aggregate gains. These results demonstrate that the proposed method improves ship recall and high-IoU oriented localization while reducing the parameter count and GFLOPs. Full article

Erbium-doped fibers (EDFs) suffer from radiation-induced absorption (RIA) and radiation-induced gain variation (RIGV) under ionizing radiation, which limit their applications in space optical communication systems. To address this issue, a novel core–cladding Ce co-doped fiber is proposed, in which core Ce suppresses radiation-induced defect formation while cladding Ce reduces localized energy deposition in the fiber core. GEANT4 Monte Carlo simulations were performed to optimize the Ce-doping configuration, and three kinds of fiber samples, namely conventional erbium-doped fiber (EDF), core Ce co-doped EDF (CEDF1), and core–cladding Ce co-doped EDF (CEDF2), were fabricated for experimental validation. At a total dose of 1200 Gy, CEDF2 exhibits a 40.5% reduction in RIA and a 58.1% reduction in RIGV compared with the conventional EDF. The experimental results demonstrate that the proposed core–cladding Ce co-doped structure effectively enhances the radiation resistance of erbium-doped fibers and provides a practical design strategy for radiation-hardened active optical fibers. Full article

Forests play an important role in timber and fiber production, carbon storage, biodiversity conservation, and various other ecosystem services, necessitating accurate and scalable inventory methods. Recent advances in remote sensing have enabled large-scale forest monitoring; however, challenges remain in extracting reliable information across varying spatial, temporal, and environmental conditions. Deep learning has emerged as a promising tool for addressing these limitations by learning complex patterns from diverse remote sensing data sources. This review synthesizes deep learning applications in forest inventory methods across three tasks: tree counting and localization, tree species identification, and tree measurement. In total, we evaluated 122 unique primary studies (37 for tree counting and localization, 57 for species identification, and 29 for tree measurement, with one study contributing to both the counting/localization and measurement tasks) spanning terrestrial, unmanned aerial vehicle (UAV), airborne, and satellite platforms, with a primary focus on optical imagery, Light Detection and Ranging (LiDAR) data, and their fusion. Across these studies, deep learning models frequently outperformed conventional machine learning and statistical baselines, with reported gains including up to 18% improvements in biomass estimation accuracy from data fusion and individual-tree species classification accuracies exceeding 90% for select architectures. However, performance differences were influenced strongly by forest structure, species complexity, sensor capability, and validation design. Counting and localization were generally more reliable in plantations than in complex natural or urban forests, while LiDAR was particularly valuable in dense, multilayer canopies. Species-identification accuracy was highest in studies with small, distinctive species sets, whereas mixed stands with many species showed lower accuracy. Only about a third of the reviewed studies (42 of 122) were externally validated on data or sites independent of model training, and reference data for tree measurement tasks were rarely based on direct destructive sampling. External validation often revealed lower performance than within-study testing, suggesting that reported accuracies may overestimate performance in new locations or conditions. Major advances are evident in the growing use of high-resolution UAV and smartphone-based imagery for tree-level analysis, the continued value of LiDAR for structural characterization, and the increasing integration of multimodal data fusion to improve detection, classification, and measurement accuracy. Persistent challenges include the limited availability of high-quality reference data, class imbalance and inconsistent species coverage, and weak model transferability across forest types, environmental conditions, and geographic regions. Future progress will likely depend on three priorities: development of larger and more standardized labeled datasets, stronger integration of structural, spectral, and phenological information, and the design of more transferable and application-oriented deep learning frameworks. Overall, this review provides a comprehensive, quantitatively grounded overview of deep learning-driven forest inventory methods and outlines future directions for improving scalability and applicability in forest monitoring and management. Full article

Marine organisms from deep-sea environments have attracted considerable attention in drug discovery because they produce structurally diverse natural products. However, Norwegian deep-sea ecosystems remain largely unexplored in terms of natural product chemistry and bioactivity. In this study, a deep-sea sponge belonging to the genus Phakellia was selectively collected using a low-impact, remotely operated vehicle approach from the Mohn’s Treasure area in the Norwegian Sea at a depth of 2858 m, representing one of the deepest sponge samples investigated in Norwegian waters to date. Chemical investigation of this specimen resulted in the isolation of three new glycosylated fatty acid amides, phakelliosides A–C (1–3), together with two known compounds, 11-(S)-myxillin B (4) and 11-(S)-myxillin C (5), which were isolated as pure individual compounds from natural sources for the first time, with their absolute configurations unambiguously established. These compounds were isolated using a combination of chromatographic techniques, and the structures of the isolated compounds were elucidated by comprehensive spectroscopic analyses, including 1D and 2D NMR and UHPLC–HRMS data. The absolute configurations were determined by optical rotation analysis following acid hydrolysis. Compounds 1–5 were subjected to preliminary antibacterial screening against Enterococcus faecalis, Staphylococcus aureus, Streptococcus agalactiae, Escherichia coli, and Pseudomonas aeruginosa at 100 µg/mL. Under the screening conditions, compound 3 produced the lowest OD₆₀₀ value against E. faecalis (OD₆₀₀ = 0.154), whereas generally higher OD₆₀₀ values were observed against the Gram-negative strains. To the best of our knowledge, this is the first report describing the preliminary antibacterial screening of glycosylated fatty acid amides isolated from Norwegian deep-sea organisms. These findings expand current knowledge of the chemical diversity of natural products associated with Norwegian deep-sea sponges. The biosynthetic origin of these metabolites remains unresolved and warrants further investigation. Full article

The operational satellite monitoring of pasture biomass requires models that transfer beyond the properties on which they were calibrated. We present a hierarchical, open-source cascade that scales in situ clip-and-weigh biomass (n = 1120 samples across eleven sites on five Queensland properties) through UAV digital-surface-model imagery to Sentinel-2 predictions, using TabPFN—a pre-trained transformer foundation model for small tabular data—as the regressor at all three nested spatial scales. Under a leave-one-site-out (LOSO) protocol on twenty site–date aggregates across nine sites, spectral-only Sentinel-2 models failed to transfer (best $R^2=-0.15$, RMSE 4.62 t ha⁻¹). Appending open climate (Open-Meteo ERA5) and topsoil (SoilGrids 2.0) covariates and evaluating five learners (GBM, RF, XGBoost, TabPFN, and GBM + TabPFN stack) on log-transformed biomass increased LOSO $R^2$ to $-0.05$ and reduced RMSE to 4.43 t ha⁻¹; a leaf-nitrogen growth trajectory predicted by the TabPFN nitrogen regressor from our earlier pasture-chemistry work reduced pixel-level LOSO RMSE by a further 4%. Three alternative covariate classes—BARRA-R2 reanalysis climate, three independent fractional-cover products, and Sentinel-1 C-band SAR backscatter—were tested and rejected, all hitting the same RMSE floor. The symmetric negative results indicate that the residual LOSO ceiling on the current nine-property footprint is a sample-size and optical-saturation limit rather than a feature-engineering one; the most tractable operational path forward is to stratify the production model by climatic zone and Queensland Land Type rather than pursuing further covariates within a single global learner. Full article

Cocaine-induced ocular toxicity is an increasingly acknowledged but often underestimated clinical condition that includes a wide range of eye-related symptoms. The extensive recreational use of cocaine, together with its powerful sympathomimetic and vasoconstrictive effects, leads to various ocular problems that can impact all anatomical components of the eye. This narrative review aims to deliver a thorough and current synthesis of the existing research on the epidemiology, pathophysiology, clinical symptoms, and therapy of cocaine-related ocular illness. Evidence suggests that cocaine-induced ocular damage involves vascular, local toxic, neuronal, and immunological processes. Reported manifestations in the literature include the ocular surface, optic nerve, posterior segment, orbit, and adnexal tissues. Novel imaging techniques have revealed subclinical retinal microvascular changes in chronic users, suggesting a continuum from initial vascular dysregulation to manifest ischemic injury. Notwithstanding increased awareness, the existing evidence is constrained by heterogeneity, small sample sizes, and a prevalence of case reports. A comprehensive understanding of the pathophysiological mechanisms and long-term ocular effects is crucial for enhancing diagnostic precision, directing management, and informing preventive measures. Enhanced multidisciplinary collaboration between ophthalmologists and addiction experts is essential to tackle this intricate and dynamic clinical dilemma. Full article

This study addresses a major challenge in hyperthermia therapy: achieving fast and efficient heat generation in target tissues. A novel magneto-plasmonic nanostructure is developed by combining gold nanorods (GNRs) and iron oxide nanoparticles (IONPs), each synthesized independently to compare their individual and combined heating performance. A key innovation was the use of (3-mercaptopropyl)trimethoxysilane (MPTMS) as a covalent linker, enabling the stable integration of both components into a single hybrid system. Under simultaneous near-infrared (NIR) laser and alternating magnetic field exposure, the hybrid nanostructure exhibited a rapid and intense temperature rise, surpassing the effects of either material alone. A control sample consisting of a physical mixture of the two components (GNR + IONP) reached SAR values comparable to those of the covalently linked hybrid (GNR + MPTMS + IONP) under simultaneous excitation, the two being equal within experimental error. Although the chemical linkage modifies the optical absorbance of the GNRs and may partially restrict the Brownian relaxation of the magnetic nanoparticles, these effects do not translate into a measurable loss of heating efficiency under combined activation. Importantly, the covalent linkage yields a robust, structurally stable assembly whose components move together—an essential requirement for functionalities such as magnetic guidance and targeted delivery of the whole nanostructure, which a simple physical mixture cannot provide. As a physicochemical proof of concept, these results establish the dual-mode heating performance of the hybrid nanostructure in aqueous suspension and motivate the biological evaluation required for any future therapeutic use. Full article

This paper investigates the feasibility of retrieving stratospheric aerosol extinction fields from GEO–LEO limb measurements using a dedicated inversion framework. The retrieval problem is severely ill-posed and involves a fundamentally three-dimensional observation geometry. To obtain stable solutions, variations of the extinction field in the cross-track direction are neglected, reducing the problem to the retrieval of radial and weakly horizontally varying aerosol distributions within a quasi-planar GEO–LEO geometry. Two simplified retrieval strategies are considered. In the first strategy, the radial extinction profile and the horizontal extent of the aerosol field are retrieved using external aerosol optical thickness observations as additional constraints. In the second strategy, the extinction field is represented by a parametric model and the corresponding model parameters are retrieved. To reduce the computational complexity, a simplified single-scattering forward model is adopted. The inversion problem is formulated as the minimization of a regularized Tikhonov function and is solved using a multistart optimization framework combining global random sampling, validation and selection of admissible starting points, local bounded optimization, discrepancy-principle filtering, and clustering of candidate solutions. Numerical simulations for a broad range of synthetic aerosol scenarios show that the proposed methodology is capable of reproducing the dominant aerosol structures with good accuracy. Although the retrieval problem remains strongly ill-conditioned, the effective degrees of freedom indicate that GEO–LEO limb measurements contain substantial independent information about the aerosol field and provide a promising basis for retrieving both radial aerosol extinction profiles and aspects of their horizontal structure. Full article