Search Results (20,948)

Rapid urbanization necessitates high-frequency monitoring of construction-ready bare land to timely detect and prevent illegal construction. However, the utility of optical imagery is often compromised in cloud-prone regions. While Synthetic Aperture Radar (SAR) offers all-weather capabilities, it struggles to distinguish construction-ready bare land from recently harvested agricultural land, leading to severe false alarms. To address the conflict between high-frequency monitoring and semantic identification, this study proposes the SAR-based Change Screening and Optical-Scene-Informed Identification (SCS-OI) framework. The first stage performs high-recall candidate screening based on SAR backscattering changes, while the second stage incorporates historical cloud-free optical imagery as semantic guidance, enabling refined identification without requiring synchronous optical data. Experiments in Guangzhou demonstrate that the framework achieves a False Alarm Rate of 13.31%, Recall of 90.63%, Precision of 74.81%, F1-score of 81.95%, and IoU of 69.43%. Compared with the SAR-only baseline (FR = 22.4%), the two-stage design reduces false alarms while maintaining high recall. Other deep learning baselines exhibit lower F1-scores (59–73%), highlighting the effectiveness of the overall framework. These results show that the proposed two-stage framework effectively integrates high-recall candidate screening and semantic-guided refinement, providing a robust solution for high-frequency monitoring of construction-ready bare land in cloud-prone regions of Guangzhou. Full article

Pig behavior statistics can reflect their health status. Conventional approaches depend on manual observation to derive behavioral information from video recordings, a process that demands substantial time and human effort. To overcome these limitations in indoor intensive farming environments, this study introduces an effective approach for recognizing pig behaviors, employing an enhanced YOLOv8n architecture. The approach utilizes advanced object detection algorithms to automatically identify pig behaviors, including stand, lie, eat, fight, and tail-bite, from overhead video footage of the enclosure. First, images of daily pig behaviors are collected using cameras to build a pig behavior dataset. To boost detection accuracy, the SE attention mechanism is embedded within the feature extraction backbone of the YOLOv8n network to enhance its representational capacity, strengthening the model’s capacity to grasp overarching contextual information and improve the expressiveness of extracted features. The GIoU loss function is employed during training to reduce computational cost and accelerate model convergence. Moreover, integrating Ghost convolution into the backbone significantly reduces both computational complexity and the total number of parameters. The experimental findings reveal that the optimized YOLOv8n model contains just 1.71 million parameters, marking a 42.93% reduction relative to the baseline model. Its floating-point operations total 5.0 billion, indicating a 38.27% decrease, while the mean average precision (mAP@50) reaches 96.8%, surpassing the original by 2.6 percentage points. Compared with other widely used YOLO-based object detection frameworks, the proposed approach achieves notably higher accuracy while requiring significantly lower computational resources and model complexity. Full article

Coastal freak waves (CFWs) are sudden and hazardous wave events that occur near shorelines and can pose serious threats to coastal visitors and infrastructure. Due to the complex interactions among coastal bathymetry, wave dynamics, and environmental conditions, the mechanisms governing CFW formation remain poorly understood, making reliable prediction difficult. This study investigates the feasibility of applying machine learning techniques to predict CFW occurrences using observational environmental data. Three machine learning algorithms, the Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN), were developed to generate probability-based predictions of CFW events. Environmental variables derived from buoy observations, including wave characteristics, wind conditions, swell parameters, wave grouping indicators, and nonlinear wave interaction indices, were used as model inputs. Hyperparameters were optimized using grid search combined with k-fold cross-validation. The results show that all three models achieved comparable predictive performance, with AUC values close to 0.80 and overall prediction accuracy around 74%. The ANN model achieved the highest recall, indicating strong capability in detecting CFW events, while the RF and SVM models showed more balanced precision and recall. Analysis of high-probability prediction events suggests that CFW occurrences are associated with swell-dominated conditions, strong wave grouping behavior, and enhanced nonlinear wave interactions. These results demonstrate that machine learning provides a promising framework for probabilistic prediction of coastal freak waves and has potential applications in coastal hazard assessment and early warning systems. Full article

Citrus trees exhibit severe alternate bearing, resulting in significant annual yield fluctuations and posing substantial challenges to orchard management planning. Accurate citrus bud counting provides an effective solution by supplying essential data for tree-level and orchard-level yield prediction. However, citrus buds are extremely small (5–10 mm in diameter) and are frequently occluded by leaves during the flowering stage, which makes precise detection highly challenging in complex orchard environments. To address these challenges, this paper proposes a Contextual and Frequency-Aware Detector (CFADet) for robust citrus bud detection. Specifically, an Enhanced Feature Fusion (EFF) module is introduced in the neck to refine multi-scale feature aggregation and strengthen information flow for small targets. A Contextual Boundary Enhancement Module (CBEM) is designed to capture surrounding contextual cues and enhance boundary representation through dimensional interaction and max-pooling operations. To suppress background interference, a Frequency-Aware Module (FAM) is developed to adaptively recalibrate frequency components in the amplitude spectrum, thereby enhancing target features while reducing background noise. In addition, Spatial-to-Depth Convolution (SPDConv) is employed to reconstruct the backbone to preserve fine-grained bud features while reducing model parameters. Experimental results show that CFADet achieves 81.1% precision, 80.9% recall, 81.0% F1-score, and 87.8% mAP, with stable real-time performance on mobile devices in practical orchard scenarios. This study presents a preliminary investigation into robust citrus bud detection in real-world orchard environments and provides a promising technical foundation for intelligent orchard monitoring and early yield estimation, while further validation on larger and more diverse datasets is still required. Full article

Anti-analysis techniques, also known as evasive techniques, enable Windows malware to detect and evade dynamic inspection environments, undermining the effectiveness of virtual-machine and sandbox-based inspection. Despite extensive prior research, no unified classification has been paired with a large-scale empirical evaluation of static detection capabilities for these behaviors. This paper addresses this gap by presenting a comprehensive classification and detection framework. We consolidate 94 anti-analysis techniques from academic, community, and threat-intelligence sources into nine mechanistic categories and derive corresponding YARA rules for static identification. In total, 82 YARA signatures were authored or refined and evaluated on 459,508 malware and 92,508 goodware samples. After iterative refinement using precision thresholds, 42 rules achieved high accuracy (≥75%), 16 showed moderate precision (50–75%), and 24 were discarded due to unreliability. The results indicate strong static detectability for firmware- and BIOS-based checks, but limited precision for timing-based evasions, which frequently overlap with benign behavior. Although YARA provides broad coverage of observable artifacts, its static nature limits detection under obfuscation or runtime mutation; our measurements therefore represent conservative estimates of technique prevalence. All validated rules are released in an open-source repository to support reproducibility, improve incident-response workflows, and strengthen prevention and mitigation against real-world threats. Future work will explore hybrid validation, container-evasion extensions, and forensic attribution based on signature co-occurrence patterns. Full article

To meet daily grasping needs under lightweight, low-complexity wearable constraints, this study proposes an underactuated multi-finger prosthetic hand with transmission–control co-design to achieve predictable multi-joint synergies and stable grasps under limited actuation. The prototype uses six miniature motors to drive 14 joint degrees of freedom (DOFs): four fingers have active metacarpophalangeal actuation with tendon-driven underactuated proximal and distal interphalangeal joints, while the thumb provides two independently controlled DOFs for opposition expansion and posture adjustment. It supports five-finger power grasps, tripod pinches, and lateral pinches. To mitigate tendon slack and stroke inconsistency, active/passive tendon-length constraints are defined, and an equal-stroke configuration is obtained via chord-to-arc mapping. A layered STM32F767-based controller combines a reference rotation range limit (free motion) with encoder speed-decay detection (contact/near-stall) to realize per-finger termination and overdrive protection without force/tactile sensors. Experiments report a total mass of 176.6 g and a peak single-finger driving force of approximately 2.8 N. Following the Feix GRASP taxonomy (33 types), the hand reproduces 24 types (72.7%), covering power, intermediate and precision grasps, both thumb abduction/adduction postures, and palm–pad–side opposition/contact, with stable grasp formation across objects of varying geometries. Full article

Salmonella enteritidis (SE) is recognized as a primary etiological agent of foodborne infection and food poisoning. Selective and sensitive determination of SE in animal-derived products is of great importance for ensuring safety in the food industry. Here, we report a highly sensitive and specific competition assay for detecting SE in eggs without interference from background fluorescence, by using persistent luminescent nanoparticles (PLNPs) as luminescent probes in combination with aptamer recognition and magnetic separation. Initially, the SE-specific aptamer (SEapt), as previously reported, was conjugated onto the surface of Fe₃O₄ magnetic nanoparticles to serve as both the recognition and separation unit. Meanwhile, the ZnGa₂O₄:Cr (PLNPs) were functionalized with the aptamer-complementary DNA (cDNA), serving as the PL signal generator. The constructed PL aptasensor is composed of the aptamer-conjugated MNPs (MNPs-SEapt) and cDNA-functionalized PLNPs (PLNPs-cDNA), integrating the merits of the long-lasting luminescence of PLNPs, the magnetic separation ability of MNPs and the selectivity of the aptamer. This integration offers a promising approach for autofluorescence-free determination of SE in food samples. The proposed aptasensor exhibited excellent linearity in the range from 1.0 × 10²–1.0 × 10⁷ CFU mL⁻¹ with a limit of detection as low as 32 CFU mL⁻¹. The precision for 11 replicate determinations of 1.0 × 10³ CFU mL⁻¹ SE was 3.4% (relative standard deviation). The developed aptasensor achieved recoveries ranging from 98.8% to 102.8% for the determination of SE in the presence of common foodborne bacterial interferents. The method was successfully applied to the analysis of Salmonella genus in egg samples. In principle, the proposed platform may be adapted to other food matrices by substituting the target-specific aptamer, pending target-dependent optimization and validation. Full article

Precise measurement of particle concentration in dilute gas–solid two-phase flows is challenging due to low particle loading and stochastic particle motion, which lead to weak signals and detection blind zones. This study develops a swirling-flow-enhanced triboelectric nanogenerator (SF-TENG) using active flow field regulation to enhance particle–wall interactions. Through CFD optimization of guide vane geometry, the SF-TENG achieved a nearly twenty-fold increase in short-circuit current compared to non-swirling configurations. The swirling flow exhibited a particle-size-dependent enhancement mechanism. For fine particles, the improvement was mainly attributed to an increased collision ratio. For coarse particles, it resulted from enhanced charge transfer per single impact. The swirling flow continuously improved the reliability and sensitivity of detection across all particle sizes. These findings provide valuable insights for designing highly sensitive, self-powered flow meters with minimized blind zones for gas–solid monitoring. Full article

The aim of this study was to develop an effective and fast tool to support the automatic segmentation of skin lesions, with particular emphasis on the precise differentiation between malignant and benign lesions. In response to the problem of high false positive rates in existing CAD systems, modern neural network architectures from the YOLO family (YOLOv8, YOLOv9, YOLOv11, YOLOv12, and YOLOv26) were used in this research. The models were trained and evaluated on a new, balanced dataset (7000 images) based on the ISIC archive, where the key innovation was the introduction of a dedicated background class representing healthy skin. Through a multi-stage, rigorous optimization process, it was demonstrated that the yolov11s-seg model is highly effective for this task. It achieved a strong balance between effectiveness and processing speed, obtaining an mAP@50 score of 0.840 and an overall precision of 0.852. From a clinical perspective, the model’s high sensitivity (85.9%) in detecting the most aggressive lesion, invasive melanoma (MI), is particularly noteworthy. Thanks to its extremely short inference time (only 4.8 ms), the proposed yolov11s-seg variant overcomes the limitations of heavy hybrid architecture, providing a stable and highly efficient solution showing significant potential for deployment in real-time medical mobile applications. Full article

Large-sized, room-temperature ultrafast scintillator single crystals are highly demanded for fast timing applications such as time of flight–positron emission tomography, high-speed medical imaging, and pulse heavy-ray detection. Sub-nanosecond scintillation was discovered in 16 mm sized CsPbCl₃Br_x single crystals in our previous research. In this work, the crystal size of CsPbCl₃Br_0.03 was enlarged to 2 inches (50.8 mm). Meanwhile, by precisely optimizing the vertical Bridgman growth process, we further increased the concentration of Br dopant to realize even faster scintillation decay. In this study, we conducted a series of tests on the grown crystals, including temperature-dependent photoluminescence tests, alpha particle excitation tests, X-ray imaging tests, etc. Via the strategy of the incorporation of Br₂, Br dopant introduces highly efficient fast recombination centers in perovskite CsPbCl₃Br_0.03 crystals, resulting in an unprecedently fast scintillation decay time of 303 ps under ²⁴¹Am α-particle excitation, which is significantly shorter than that of the pure CsPbCl₃ and all other perovskites by at least two orders of magnitude. Benefiting from the excellent optical transparency and high crystalline quality of the CsPbCl₃Br_0.03 crystal, an X-ray spatial resolution of up to 20 lp/mm is achieved. These results further demonstrate the great potential of large-sized CsPbCl₃Br_x single crystals for fast timing applications. Full article

During the interpretation of Borehole Radar (BHR) B-scan profiles, the accurate determination of the azimuth of geological targets in three-dimensional space is a critical issue for achieving precise anomaly localization and spatial structure inversion. However, existing directional BHR anomaly localization methods exhibit limited intelligence, insufficient adaptability to multi-site data, and weak generalization capability, rendering them inadequate for engineering applications under complex geological conditions. To address these challenges, a robust deep learning model, termed BSS-Pose-BHR, is developed based on YOLOv11n-pose for keypoint detection in directional BHR profiles. The model incorporates three key optimizations: Bi-Level Routing Attention (BRA) replaces Multi-Head Self-Attention (MHSA) in the backbone to improve computational efficiency; Conv_SAMWS enhances keypoint-related feature weighting in the backbone and neck; and Spatial and Channel Reconstruction Convolution (SCConv) is integrated into the detection head to reduce redundancy and strengthen local feature extraction, thereby improving suitability for keypoint detection tasks. In addition, a three-dimensional electromagnetic model of limestone containing a certain density of clay particles is established to construct a simulation dataset. On the simulated test set, compared with current mainstream deep learning approaches and conventional directional borehole radar anomaly localization algorithms, BSS-Pose-BHR achieves superior performance, with an mAP50(B) of 0.9686, an mAP50–95(B) of 0.7712, an mAP50(P) of 0.9951, and an mAP50–95(P) of 0.9952. Ablation experiments demonstrate that each proposed module contributes significantly to performance improvement. Compared with the baseline, BSS-Pose-BHR improves mAP50(B) by 5.39% and mAP50(P) by 0.86%, while increasing model weight by only 1.05 MB, thereby achieving a reasonable trade-off between detection accuracy and complexity. Furthermore, indoor physical model experiments validate the effectiveness of the method on measured data. Robustness experiments under different Peak Signal-to-Noise Ratio (PSNR) conditions and varying missing-trace rates indicate that BSS-Pose-BHR maintains high detection accuracy under moderate noise and data loss, demonstrating strong engineering applicability and practical value. Full article

Articular fractures require precise anatomical reduction and rigid fixation to heal appropriately. In veterinary cases that involve fracturing of the lateral humeral condyle, cortical bone screws inserted in lag fashion with Kirschner wire are the preferred method for surgical fixation. However, relatively high complication rates associated with cortical lag screws (CLSs) highlights the need to investigate alternate screw designs. Variable-pitch headless compression screws (VPHCSs) are unique as they advance beneath the cortical surface. Although the use of VPHCSs are widely utilised in human orthopaedics, the current use in veterinary orthopaedics is limited. This study aimed to evaluate the peak interfragmentary force (PIF) and area of compression (AOC) generated by a 3.5 mm self-tapping cortical screw placed in lag fashion and a 3.5 mm VPHCS inserted to four depths. PIF and AOC were measured using a pressure-sensitive film placed between two blocks of polyurethane foam (0.24 g/cm³), simulating a transverse fracture. CLSs were inserted by hand into predrilled 2.5 mm pilot holes. PIF and AOC were measured at full insertion. VPHCSs were placed into predrilled 2.5 mm pilot holes, followed by a 3.5 mm tapered countersink. The screw was inserted until the head was level with the surface. PIF and AOC were measured between the two blocks. The screw was continued until the head was at a depth of 2, 5, and 9 mm below the surface, and the PIF and AOC were measured again at each stage. There was no detectable difference in PIF and AOC between CLSs and VPHCSs countersunk to −2 mm (PIF–CLS: Mean = 12.886, SD = 2.370; 2 mm: Mean = 17.301, SD = 8.858, p = 0.319; AOC–CLS: Mean = 0.936, SD = 0.291; 2 mm: Mean = 0.925, SD = 0.447, p = 0.872). VPHCSs countersunk to −5 mm and −9 mm produced significantly greater PIF compared to CLSs (5 mm: Mean = 16.086, SD = 6.799, p = 0.002; 9 mm: Mean = 34.987, SD = 4.015, p < 0.001). VPHCSs countersunk to −5 and −9 mm produced significantly greater PIF and AOC compared to CLSs in this model. Further investigation is required to produce recommendations for clinical use. Full article

Ultra-precision thinning of hard and brittle materials like monocrystalline silicon demands high dynamic stability in thinning spindle. To address the challenge of accurately detecting subtle spindle chatter anomalies in industrial environments characterized by high noise and limited data, this paper proposes a physics-guided dual-stream attention fusion transfer network (PG-AFNet). First, a physics-guided signal preprocessing method was developed. Using variational mode decomposition (VMD) and continuous wavelet transform (CWT) masking, one-dimensional dynamic features and high-frequency regions of interest (ROIs) rich in transient impact features were extracted. Second, the PG-AFNet architecture was designed. By introducing an attention mechanism, it achieves deep integration of one-dimensional purely dynamic sequences with two-dimensional spatiotemporal visual textures to capture surface damage features caused by subtle vibrations. Finally, systematic validations were conducted using a real silicon wafer thinning dataset with 197 real samples. By overcoming small-sample limitations via physical augmentation, PG-AFNet achieved an 82.45% (86.64% after data augmentation) diagnostic accuracy, significantly outperforming traditional baselines. Furthermore, a large-scale cross-load validation on the diverse CWRU dataset yielded an exceptional 99.68% accuracy under mixed-load conditions, conclusively verifying the model’s robust domain generalization. Lastly, a rigorous ablation study explicitly quantified the indispensable contributions of the physics-guided dual-stream architecture and attention fusion. This research provides a feasible theoretical foundation for intelligent surface quality monitoring in semiconductor hard-brittle material processing. Full article

Alcoholic fermentation in winemaking is a complex bioprocess governed by physicochemical parameters such as temperature, density, pH, CO₂ and redox potential, which critically affect yeast metabolism and wine quality. This review provides an integrated analysis of fermentative dynamics and emerging sensorization technologies, highlighting how their combined implementation enables real-time monitoring and advanced control in precision enology. Advances in conventional physicochemical sensors, spectroscopic techniques (NIR/MIR/UV-Vis) and non-conventional devices (e-noses, electronic tongues) integrated into IoT platforms enable continuous data acquisition, overcoming traditional manual sampling limitations. Predictive modeling, including kinetic models, machine learning approaches (e.g., Random Forest, XGBoost) and model predictive control (MPC/NMPC), supports anomaly detection, optimization of enological interventions and energy-efficient thermal management, while virtual sensors based on Kalman filters improve the estimation of non-measurable states (e.g., biomass, ethanol kinetics). Despite current challenges in calibration and interoperability, these innovations foster sustainable and reproducible winemaking under climate variability and pave the way for digital twins and semi-autonomous fermentation systems. Full article

Background/Objectives: The co-occurrence of asthma and obesity presents a significant clinical challenge, but the underlying mechanisms remain unclear. Reduced adiponectin and vitamin D levels have been associated with both conditions, suggesting that their potential modulatory roles warrant further investigation. This study aimed to evaluate whether vitamin D and adiponectin levels differ among pediatric groups defined by their asthma and obesity status, to better characterize the metabolic and inflammatory profile of the obesityasthma phenotype. Methods: A total of 120 participants aged 6–18 were enrolled and categorized into four groups: Asthma (n = 30), Obesity (n = 30), Asthma + Obesity (n = 30), and Control group (n = 30). All participants underwent lung function testing, anthropometric assessment and measurement of fraction of exhaled nitric oxide (FeNO). Participants were further categorized according to BMI percentiles. Adiponectin levels were measured using ELISA, while vitamin D levels were detected using HPLC. Results: Vitamin D levels and lung function parameters did not differ across groups, although deficiency was most prevalent in the obesity group. FeNO was elevated in asthmatics relative to obese children (p = 0.038) and in obese asthmatics compared with both controls (p = 0.040) and obese children (p = 0.021). Adiponectin levels were lower in obese asthmatic children compared to the controls (p = 0.024). A similar difference was observed between the controls and obese asthmatics among children with low vitamin D levels (p = 0.014). Conclusions: The dominant mechanisms underlying the obesity–asthma phenotype remain unclear; however, our findings indicate a link between adiponectin dysregulation and heightened airway inflammation, as evidenced by increased FeNO levels, though the precise pathways involved are still not well-understood. The role of vitamin D appears less consistent. These results highlight the need for further research to clarify the interplay between metabolic and inflammatory pathways and to support more personalized management strategies in children with obesity-related asthma. Full article