Search Results (864)

Urban forest walkways provide accessible forest recreation opportunities within cities, but little is known about how different walkway landscape environments are associated with older adults’ perceived psychological restoration. This study examined three representative urban forest walkways in Fuzhou, China, by integrating image-based landscape classification with photo-elicitation questionnaires. A total of 1188 field photographs were collected, of which 963 valid images were used for landscape analysis. A SegNet-based semantic segmentation model was used to extract major landscape elements, and the images were subsequently classified into four landscape environment types. Questionnaire data were obtained from 415 older walkway users, generating 4980 respondent–image evaluation records. The SegNet model achieved a mean IoU of 79.4%. The high-naturalness canopy-immersive type received the highest scores for tranquillity, refuge, naturalness, landscape preference, and perceived psychological restoration, whereas the semi-open viewing-platform type showed the highest spaciousness and prospect scores. Landscape preference was positively associated with perceived psychological restoration in the mixed-effects model and partially mediated the associations between selected perceived landscape attributes and perceived psychological restoration. These findings suggest that age-friendly urban forest walkway design should maintain canopy immersion, provide semi-open viewing nodes, improve resting and interaction spaces, and refine forest-edge urban-interface sections. Full article

Staphylococcus pseudintermedius is a common opportunistic pathogen in companion animals and a leading cause of skin and soft tissue infections (SSTIs). Despite its clinical relevance, the genomic determinants underlying pathogenicity and the transition from commensal carriage to invasive infection remain poorly understood. This study aimed to identify the genomic determinants of SSTI pathogenic potential in S. pseudintermedius and to determine whether pathogenicity is driven by single, major-effect virulence genes or by polygenic genome-wide genetic architecture. Using accessory gene-based and unitig-based genome-wide association studies (GWASs), employing a linear mixed model, across a genetically diverse collection of S. pseudintermedius isolates spanning multiple phylogenetic clades, we found that disease and carriage isolates showed no phylogenetic clustering. SSTI pathogenicity exhibited high narrow-sense heritability. Surface-associated LPXTG-anchored proteins, particularly spsF, harbored the strongest associations, with additional signals in iron metabolism (narH, sufB) and oxidative stress tolerance (ahpC). Random Forest classification validated GWAS signals. SSTI pathogenicity in S. pseudintermedius reflects a polygenic architecture driven by cumulative variation across surface-associated, metabolic, and stress-response loci, shifting focus from single virulence genes to genome-wide genetic variation. Full article

High-precision mapping of tree species composition is essential for sustainable forest management, biodiversity assessment, and ecosystem monitoring. Airborne hyperspectral imagery provides rich spectral and spatial information that enables detailed species discrimination. However, traditional single-stream convolutional neural networks (CNNs) often fail to fully exploit multi-dimensional features and are susceptible to spectral redundancy. In this study, we propose the MS-CBAM-TSCNet (Multi-Stream Convolutional Block Attention Deep Neural Network), a novel architecture specifically designed for tree species classification using aerial hyperspectral data. The proposed model integrates three parallel processing streams: a 1D spectral branch for capturing reflectance signatures, a 2D spatial branch for modeling contextual patterns, and a 3D spectral–spatial branch enhanced with Convolutional Block Attention Modules (CBAMs) to adaptively recalibrate channel-wise and spatial–spectral features. An adaptive gated fusion mechanism with attention-based weighting is introduced to dynamically combine the complementary representations extracted from the three streams, improving robustness to spectral redundancy and class imbalance. The method was evaluated on an airborne HyMap hyperspectral dataset (125 bands, with 4 m spatial resolution) acquired over a mixed boreal forest in Karlsruhe, Germany, comprising five dominant tree species. Using five-fold cross-validation on an augmented dataset, the MS-CBAM-TSCNet achieved an overall accuracy of 96.8%, a Kappa coefficient of 0.96, and a macro F1-score of 0.966, outperforming conventional 1D, 2D, and 3D CNNs, as well as a hybrid CNN-SVM approach across all evaluation metrics. An ablation study further confirms the complementary contributions of the multi-stream architecture, CBAM attention, and adaptive gated fusion. Pixel-wise classification maps demonstrate improved boundary delineation and reduced misclassification in mixed stands, highlighting the effectiveness of the proposed framework for operational forest inventory and ecological monitoring. Full article

Motorcyclists are among the most vulnerable road users at roundabouts, particularly in mixed-traffic environments where rider behaviour may be influenced by geometric design. This study investigates the association between roundabout geometry and risky motorcyclist behaviour using the Chi-squared Automatic Interaction Detection (CHAID) method. Video-based observations were conducted at four selected roundabouts in Kuching, Sarawak, Malaysia, generating 15,937 risky-behaviour events from 5400 observed motorcyclists. Six risky-behaviour categories were analysed, covering entry, circulation, and exit manoeuvres. The CHAID results showed that entry radius was the primary geometric factor associated with risky motorcyclist behaviour, while exit radius and exit width acted as secondary variables under specific entry-radius conditions. Four behavioural scenarios were identified. Entry radii of 15–32 m were associated with mixed risky-behaviour patterns, with lane splitting during circulation being the most frequent behaviour. Entry radii of 37–40 m combined with exit radii ≤ 12.43 m were associated with failure to signal before exiting. Entry radii of 43–49 m combined with exit radii ≤ 25.3 m were associated with improper lane positioning when exiting. Larger entry radii of 49–64 m combined with exit widths of 7.38–9.06 m were associated with close stopping or potential blind-zone positioning. The model validation results indicated moderate internal classification performance, supporting the use of CHAID as an interpretable threshold-identification tool rather than a high-precision predictive model. The findings demonstrate that risky motorcyclist behaviour at roundabouts is shaped by non-linear interactions between entry and exit geometric elements. From a sustainability perspective, these results provide preliminary evidence for behaviour-sensitive roundabout design, safety assessment, and policy-oriented geometric improvements that support safer and more inclusive urban transport systems in motorcycle-dominant mixed-traffic contexts. Full article

The increasing realism of deepfake videos has intensified the need for reliable video-level detection systems, but benchmark performance must be interpreted together with generalization, temporal-modeling, and efficiency limits. This study presents a reproducible multi-backbone framework for detecting manipulated videos on the official Celeb-DF v2 benchmark. Five pretrained image-classification architectures—ResNet-50, EfficientNet-B4, ConvNeXt-Small, ViT-Base, and Swin-Base—are fine-tuned under a unified protocol using uniform frame sampling, class-balanced training, RandAugment, MixUp, CutMix, random erasing, label smoothing, AdamW optimization, cosine learning-rate scheduling, and exponential moving average weights. During inference, frame-level fake probabilities are stabilized using horizontal-flip test-time augmentation and aggregated into video-level predictions by mean probability pooling. This aggregation is a fixed probability-pooling rule rather than an explicit temporal model. A probability-level ensemble of ResNet-50, ConvNeXt-Small, and Swin-Base combines convolutional and attention-based representations. On the official 518-video Celeb-DF v2 test set, the top-three ensemble achieves a video-level AUC of 99.967% and an average precision of 99.983%, while ResNet-50 provides the strongest single-model accuracy–efficiency trade-off. Additional analyses examine frame-to-video aggregation, ROC and precision–recall behavior, probability distributions, frame-probability stability over sampled frames, a proof-of-concept temporal-splice sensitivity test, and inference efficiency. The results demonstrate highly competitive in-dataset performance on Celeb-DF v2 only. Because no external benchmark testing, learned temporal baseline, confidence-gated cascade, or broad partial-manipulation benchmark is included, the results should not be interpreted as evidence of cross-dataset robustness, in-the-wild deployment readiness, learned temporal reasoning, or general localization capability. Cross-dataset evaluation on FaceForensics++, DFDC, WildDeepfake, and related benchmarks, probability calibration, false-positive control, explicit temporal modeling, cascade-based inference, and expanded localization evaluation are identified as priority future-work directions. Full article

This study proposes a model-based method for real-time monitoring of the lubrication regimes in dynamically loaded radial sliding bearings using Physics-Informed Neural Networks (PINN). The proposed method replaces computationally intensive elastohydrodynamic lubrication (EHD) simulations with a PINN-based surrogate model. The model predicts hydrodynamic pressure and lubricant film-thickness distributions with comparable accuracy under dynamically varying operating conditions, enabling reliable assessment of lubrication regimes. The proposed model advances the state of the art in physics-informed modelling of mixed lubrication by extending existing approaches to simultaneously account for mixed-friction regimes through the Greenwood–Tripp contact model, transient operating conditions, and bearing surface deformation. Using only the bearing load and shaft rotational speed as inputs, the resulting hydrodynamic pressure field and corresponding lubricant film thickness can be monitored, enabling the direct assessment of the lubrication regime and potential wear risk. The proposed method is applied to a validated EHD model of a 30 mm sliding bearing test rig, where EHD simulation results are used to train, validate, and evaluate the model. The proposed framework achieved an average lubricant film-thickness prediction error of 2.34% and lubrication-regime classification errors of 7.8% and 8.2% for the static and dynamic validation cases, respectively. Furthermore, the computation time for the complete 18-time-step load case was reduced from approximately 35 h to 61.2 ms. Full article

While existing AI-based microplastic monitoring studies predominantly rely on task-specific fine-tuned models, this study evaluates whether general-purpose multimodal foundation models with no spectroscopic instrumentation, minimal fine-tuning, and minimal computational resources can serve as accessible, low-cost tools for polymer classification under ecologically realistic field conditions with samples of plastic debris collected from coastal Sousse, Tunisia, a Mediterranean region experiencing anthropogenic pollution pressures. A curated dataset of 1080 high-resolution images was developed, representing six polymer groups (HDPE, LDPE, PA, PET, PP, PS, and mixed plastics). Fragments were imaged under standardized lighting conditions against natural sand backgrounds to preserve environmental realism. Each image was manually annotated using polygon-based boundaries to generate pixel-level segmentation masks and associated class labels, providing expert-validated ground truth for quantitative evaluation. Multimodal LLMs were evaluated using a composite scoring framework. Spatial accuracy was assessed using mean Intersection-over-Union (mIoU) against expert annotations, while polymer classification performance was measured using macro-averaged F1 scores across all categories. Model reliability was further evaluated through prompt stability testing and robustness analyses under controlled environmental perturbations designed to assess consistency across varying coastal imaging conditions. Results indicate that multimodal foundation models can distinguish plastic fragments from sand backgrounds, although performance varied across polymer classes and environmental perturbations. The weighted composite framework provides a structured approach for comparing model performance according to ecological monitoring objectives rather than computational metrics. These findings contribute to understanding the potential utility and current limitations of AI-based approaches for marine microplastic analysis and provide insights into coastal pollution patterns and ecosystem health within a One Health monitoring context. Full article

The precision of oil palm fresh fruit bunch (FFB) ripeness classification directly determines the extraction efficiency and chemical quality of the resulting crude palm oil (CPO). Traditional manual inspection at reception ramps remains labour-intensive, time-consuming, and subjective. To address these issues, this study presents a lightweight deep learning framework for automated oil palm FFB ripeness classification trained on a field-collected dataset of 857 images covering four ripeness classes (Over-ripe, Ripe, Under-ripe, and Unripe). Six lightweight classification backbones are evaluated, including MobileNetV2, EfficientNetV2B0/B1, and YOLO variants. An intra-class half-mixing augmentation with geometric transformation is proposed to address minority-class imbalance. Overall, YOLOv8n-cls achieved the highest accuracy (95.6%), followed by EfficientNetV2B0/B1, YOLO11n-cls, YOLO26n-cls, and MobileNetV2, respectively. In addition, all YOLO-family models achieved a recall score of 1.00 for the minority class while obtaining the highest $F_1$ scores for the other classes. The experimental results suggest that the proposed augmentation method enables lightweight deep learning models to achieve promising classification performance on an imbalanced field-collected FFB dataset while improving minority-class detection. Full article

Academic acceleration decisions for gifted students require a shift from single-scale, grade-based approaches toward integrated, multi-domain assessments that consider cognitive abilities, academic performance, and psychosocial readiness together. This mixed-methods case study examines three Saudi students who experienced academic acceleration and who present distinct cognitive–academic profiles—designated as “Outstanding” (O.F.), “Inconsistent” (O.L.), and “Inconsistent and Regressing” (O.M.)—to explore how the congruence among cognitive abilities (measured by CogAT), academic performance (evaluated through predictive descriptive modeling), and psychosocial readiness (assessed by the modified PDAS) influences acceleration decision making. Data were collected from cognitive tests, academic records, PDAS scores, and semi-structured interviews with students, teachers, and parents. Findings indicate that strong convergence across the three domains was associated with successful whole-grade acceleration, whereas domain-specific mismatch and significant weakness in fluid reasoning were associated with academic decline, psychological distress, and social difficulties. Composite scores masked substantial variations within these asymmetric patterns, leading to potentially inappropriate classifications. Although parental support varied across cases and appeared associated with student outcomes, a causal relationship could not be established within this exploratory design. The study concludes that acceleration policies should move beyond general averages toward individualized, profile-based assessments that consider subject-specific acceleration and psychosocial support when needed. All findings are preliminary observations that require replication with larger samples and longitudinal designs before definitive conclusions can be drawn. Full article

Ground-penetrating radar (GPR) and electrical resistivity tomography (ERT) are physically complementary non-destructive evaluation methods for reinforced concrete, yet their integrated diagnostic use has been limited by the absence of controlled, ground-truth-validated calibration of the joint-signature space. This paper presents a laboratory calibration programme in which a single C30/37 reinforced concrete beam (3000 mm × 300 mm × 200 mm, three T12 bars at 35 mm cover, CEM I 42.5N, w/c = 0.50) was sequentially conditioned through four controlled deterioration states—intact reference (Model A), water-filled saw-cut crack (Model B), full saturation by seven-day top-surface ponding (Model C), and chloride-induced active corrosion (Model D). Seven RES2DINV inverted ERT sections at three electrode spacings (a = 7, 15, and 30 mm) and three 800 MHz GPR profiles were acquired across the four known ground-truth conditions. The intact-reference resistivity ρ₀ = 558 Ω·m (full-section median of the mlab dataset at a = 7 mm) and GPR-calibrated velocity v = 0.095 ± 0.008 m/ns (from hyperbola fitting at 35 mm rebar cover) establish the absolute baselines. The four conditions produce systematically distinct joint signatures: Model A exhibits uniform high resistivity with clean rebar hyperbolae and no anomalous reflections; Model B produces a localised ERT low-ρ anomaly (ρ_min = 1.46 Ω·m) co-located with a negative-polarity (R = −0.68) GPR crack-mouth reflection confirming water-fill; Model C produces pervasive low-ρ with a smooth depth gradient and 50–65% GPR amplitude attenuation (−6.0 to −9.1 dB); Model D produces the same bulk GPR signatures as Model C but with a critically different ERT spatial texture—a heterogeneous near-surface layer above a sharp boundary at z ≈ 40 mm with depth-preferential low-ρ concentrated at rebar level. This depth-preferential signature, quantified here by a reproducible Depth-Preferential Index (DPI), is the primary ERT-only diagnostic criterion distinguishing active corrosion from pervasive saturation. For the Model C versus Model D distinction, the GPR response is non-discriminating; this high-risk distinction is resolved exclusively by the ERT depth-preferential criterion. The calibration demonstrates that GPR and ERT are physically non-redundant in the strict sense: neither method alone can unambiguously discriminate all four states, but their combination yields correct classification within the controlled laboratory conditions and subject to the stated qualification conditions. The corrosion state was confirmed at the regime level (chloride above the depassivation threshold, under accelerated polarisation) but was not quantified electrochemically, so the depth-preferential signature is interpreted as an indirect spatial proxy for active corrosion rather than a measurement of corrosion rate. Seven failure modes are quantitatively characterised and embedded in the framework as a priori qualification conditions. The calibrated reference values (ρ₀, A₀, Stage 2 thresholds, depth-preferential criterion) are specific to the laboratory mix and curing history and require local Stage 1 recalibration for field application. Full article

Background/Objectives: Cryptosporidiosis, caused by Cryptosporidium parvum, is a significant cause of diarrheal disease, particularly affecting young children and immunocompromised individuals. With current treatments offering limited efficacy, there is an urgent need for novel therapeutic targets. Methods: In this study, we report the cloning, expression, and functional characterization of a glutathione transferase (GST) from C. parvum (CpGST). Results: Biocomputing analysis revealed a single gene encoding a cytosolic enzyme with distinct structural features, compared to human cytosolic homologs. Structural modeling indicated a non-canonical thioredoxin fold and a truncated C-terminal domain, suggesting functional divergence. CpGST was expressed in Escherichia coli, and its enzymatic properties were characterized. Although the enzyme displayed a narrow substrate spectrum, it showed a distinct substrate preference, retaining catalytic activity toward the standard GST substrates 1-chloro-2,4-dinitrobenzene (CDNB) and cumene hydroperoxide (CuOOH). Steady-state kinetic analysis revealed limited affinity for both reduced glutathione (GSH) and CDNB. Inhibition analysis identified several polyphenols and synthetic curcumin analogues as potent inhibitors, with IC₅₀ values in the low micromolar range. Kinetic analysis with the most potent inhibitor revealed a mixed-type inhibition mechanism. Conclusions: These findings support the classification of CpGST as a structurally and functionally distinct member of the GST family, likely adapted to the parasite’s physiology and metabolism. The enzyme’s divergence from human GSTs, along with its favorable druggability profile, underscores its potential as a target for anti-cryptosporidial drug development, particularly in strategies aimed at disrupting stress response and detoxification pathways. Full article

Microwave scatterometers are capable of acquiring land surface backscattering coefficients day and night under all-weather conditions, offering advantages for snow cover monitoring. However, the relatively low spatial resolution of traditional scatterometer data limits their application in the fine-scale monitoring of snow cover distribution. To improve the spatial representation of snow cover and mitigate mixed-pixel effects in complex spring snowmelt scenarios, this study proposes an adaptive bilateral filtering scatterometer image reconstruction (SIR-ABF) algorithm based on the rotating fan-beam scanning characteristics of the FengYun-3E Wind Radar (FY-3E WindRAD). The Ku-band data of FY-3E WindRAD were reconstructed from the original 10 km resolution to an enhanced resolution of 3.125 km. Furthermore, by integrating the reconstructed scatterometer backscatter with multi-source auxiliary data, an optimal feature subset was determined through a feature selection strategy that considers both feature-label correlation and inter-feature multicollinearity. Finally, the best feature subset was combined with four machine learning (ML) models for snow cover classification. The results indicate that the Support Vector Machine (SVM) achieved the best performance, yielding an Overall Accuracy (OA), Macro-F1, and Kappa coefficient (Kc) of 91.64%, 86.47%, and 0.730, respectively. Compared with the snow cover classification results derived from the original 10 km Ku-band data, the 3.125 km reconstructed data provided more detailed spatial information and better characterized fragmented snow patches and snow transition boundaries. Further comparison with existing snow cover products demonstrated the spatial consistency and continuity of the proposed classification results, highlighting the potential of high-resolution scatterometer data for fine-scale snow cover monitoring during the spring snowmelt period in Northeast China. Full article

Hyperspectral image (HSI) classification remains a challenging task due to its high-dimensional spectral characteristics and the complex spatial heterogeneity of remote sensing scenes. Although significant progress has been made with convolutional neural networks (CNNs) and Transformer methods, existing approaches often fail to adequately model fine-grained variations in spectral curves, whilst struggling to strike a good balance between preserving local texture and modelling global context. To this end, this paper proposes a spatial–spectral dual-branch network (SFE-FM) for HSI classification, which combines spectral feature enhancement with a lightweight feature fusion mechanism to improve feature representational capacity. Specifically, a Spectral Feature Enhancement (SFE) module is designed to explicitly model spectral trends through first- and second-order differential operations, thereby enhancing potential discriminative information; at the same time, a lightweight Feature Mixing (FM) module is introduced into the network to model global dependencies across channels. In the dual-branch architecture, the spatial branch and the spectral branch extract spatial texture features and spectral semantic features respectively, and these features are adaptively reweighted using a cross-branch-guided fusion strategy (CSSF) to facilitate the effective integration of the two types of information. In addition, a multi-scale attention optimisation module (MSAO) has been introduced to enhance the response in key areas and improve the robustness of feature representations. Experiments conducted on the four benchmark datasets—WHU-Hi-HanChuan, Qingyun, Salinas and Pavia University, the proposed method achieved overall classification accuracies of 99.62%, 98.70%, 99.97% and 99.81%, respectively. Whilst maintaining a relatively low number of parameters (0.631M), it delivered competitive performance, demonstrating a good balance between classification accuracy and computational efficiency. Full article

Accurate crop mapping in middle-to-high latitude agricultural basins is essential for food security, agricultural management, and sustainable land-use planning. However, crop classification in these regions remains challenging because fragmented field patterns, mixed pixels, and overlapping phenological stages often lead to severe spectral confusion among major dryland crops. To address this issue, this study developed a Phenology–Spectral Dual-Constrained Strategy (PS-DCS) by integrating agronomic knowledge with physically constrained spectral features. The proposed framework identified August as the optimal observation window based on crop phenological divergence. Wheat was first extracted using a spectral fingerprint combining the Chlorophyll Index Red Edge (CI_RE) and Redness index. Subsequently, maize and soybean were separated within the non-wheat mask using the B6 red-edge band selected through feature separability analysis. Validation based on Sentinel-2 time-series imagery and 1056 independent field samples collected in 2025 yielded an Overall Accuracy of 95.36% with a Kappa coefficient of 0.928. Compared with RF, XGBoost, and CNN models, PS-DCS maintained competitive classification performance while substantially reducing dependence on large training datasets and complex parameter tuning. Cross-year validation during 2022–2024 further demonstrated stable spatial transferability without threshold recalibration. These results indicate that translating agronomic mechanisms into physically interpretable remote sensing rules provides an effective and transparent framework for high-precision crop mapping and long-term agricultural monitoring in complex agricultural landscapes. Full article

In coastal wetlands, Tamarix chinensis often occurs as patches intermixed with Phragmites australis, Suaeda salsa, and saline–alkaline bare soil. This mixed distribution makes tamarisk prone to omission in medium-resolution remote sensing classification, while the recall of the target species is often masked by a relatively high overall accuracy. In this study, we focused on the Yellow River Delta National Nature Reserve and developed a multi-source feature set using summer 2025 Sentinel-2, Sentinel-1, and UAV/GPS data, comprising spectral, SAR, phenological, and red-edge-oriented features. To enhance the separability between tamarisk and co-occurring herbaceous vegetation, we introduced a red-edge–vegetation-cover coupling feature (RE_cov) based on their contrasting responses in the red-edge region. Within an XGBoost framework, we evaluated the marginal contribution of this feature using feature ablation, replacement, and spatial block cross-validation. The full feature set achieved an AUC of 0.8042, a recall of 0.9340, and an overall accuracy of 0.8194 on an independent test set. Ablation and replacement experiments showed that the red-edge-oriented features contributed to both model separability and tamarisk recall, and this contribution remained evident under spatial block validation. We further converted the pixel-level extraction results into local tamarisk density grades, revealing a pattern of a few clustered cores embedded within a broad low-density background. These results suggest that target-species-oriented red-edge–vegetation-cover coupling features can improve tamarisk recall while maintaining acceptable overall accuracy, providing a spatial product to support zoned patrol and management in protected coastal wetlands. Full article