Search Results (653)

Maritime ship detection faces challenges due to complex object poses, variable target scales, and background interference. This paper introduces YOLO-PFA, a novel SAR ship detection model that integrates multi-scale feature fusion and dynamic alignment. By leveraging the Bidirectional Feature Pyramid Network (BiFPN), YOLO-PFA enhances cross-scale weighted feature fusion, improving detection of objects of varying sizes. The C2f-Partial Feature Aggregation (C2f-PFA) module aggregates raw and processed features, enhancing feature extraction efficiency. Furthermore, the Dynamic Alignment Detection Head (DADH) optimizes classification and regression feature interaction, enabling dynamic collaboration. Experimental results on the iVision-MRSSD dataset demonstrate YOLO-PFA’s superiority, achieving an mAP@0.5 of 95%, outperforming YOLOv11 by 1.2% and YOLOv12 by 2.8%. This paper contributes significantly to automated maritime target detection. Full article

The detection of internal solitary waves (ISWs) in the ocean using Synthetic Aperture Radar (SAR) images is important for the safety of marine engineering structures. Based on 4120 Sentinel SAR images obtained from 2014 to 2024, an ISW dataset covering the Andaman Sea (AS), the South China Sea (SCS), the Sulu Sea (SS), and the Celebes Sea (CS) is constructed, and a deep learning dataset containing 3495 detection samples and 2476 segmentation samples is also established. Based on the YOLOv8 lightweight model, combined with an anti-interference strategy, a multi-size block detection strategy, and a post-processing repair module, an ISW detection method is proposed. This method reduces the false detection rate by 44.20 percentage points in terms of anti-interference performance. In terms of repair performance, the repair rate reaches 85.2%, and the error connection rate is less than 3.1%. The detection results of applying this method to Sentinel images in multiple sea areas show that there are significant regional differences in ISW activities in different sea areas: in the AS, ISW activities peak in the dry season of March and are mainly concentrated in the eastern and southern regions; the western part of the SS and the southern part of the CS are also the core areas of ISW activities. From the perspective of temporal characteristics, the SS maintains a relatively high ISW activity level throughout the dry season, while the CS exhibits more complex seasonal dynamic features. The lightweight detection method proposed in this study has good applicability and can provide support for marine disaster prevention work. Full article

Two-pore channel 2 (TPC2) is a member of the endolysosomal ion channel family, playing critical roles in intracellular calcium signaling and endomembrane dynamics. This review provides an in-depth analysis of TPC2, covering its structural and functional properties, physiological roles, and involvement in human diseases. We discuss current experimental approaches to studying TPC2, including heterologous expression in plant vacuoles and computational modeling strategies. Particular emphasis is placed on the structural determinants of ion permeation, with a focus on the selectivity filter and the central cavity’s influence on channel kinetics. Furthermore, we explore emerging roles of TPC2 in viral infections, particularly SARS-CoV-2, and in cancer, including melanoma progression and neoangiogenesis. The inhibitory potential of natural compounds, such as naringenin, is also examined. By offering a comprehensive overview of current knowledge and methodologies, this review underscores the potential of TPC2 as a promising pharmacological target in both infectious and neoplastic diseases. Full article

The persistent evolution of SARS-CoV-2 has led to the emergence of antigenically distinct Omicron subvariants exhibiting increased transmissibility, immune evasion, and altered pathogenicity. Among these, recent subvariants such as JN.1, KP.3.1.1, and LB.1 possess unique antigenic and virological features, underscoring the need for continued surveillance and therapeutic evaluation. As vaccines and commercial monoclonal antibodies show reduced effectiveness against these variants, the role of direct-acting antivirals, such as Nirmatrelvir, targeting conserved viral elements like the main protease inhibitor, becomes increasingly crucial. In this study, we investigated the replication kinetics, host immune responses, and therapeutic susceptibility of three recently circulating Omicron subvariants in the K18-hACE2 transgenic mouse model, using the SARS-CoV-2 parent WA1/2020 strain as a reference. Omicron subvariants exhibited a marked temporal shift in viral infection kinetics characterized by an early lung viral titer peak (~7–8 Log PFU) at 2 days post-infection (dpi), followed by a decline (1–3 Log PFU) by 4 dpi. Pulmonary cytokine and chemokine responses (GM-CSF, TNF-α, IL-1β, IL-6) showed an earlier increase in subvariant-infected mice compared to a gradual response in WA1/2020 infection. Notably, Nirmatrelvir treatment led to significant reductions in lung viral titers in subvariant-infected mice compared to WA1/2020, surpassing its efficacy against the parent strain. These findings highlight that infection with Omicron subvariants yields a broad dynamic range in viral burden with minimum variability, while retaining a prominent therapeutic response to Nirmatrelvir. This study provides insights into the emerging subvariants’ pathogenesis and therapeutic responsiveness, reinforcing the importance of continued variant monitoring and the development of effective countermeasures. Full article

Background: The continuous evolution of SARS-CoV-2 necessitates the development of targeted strategies based on the immunological profiles of distinct age groups. Despite this imperative, comprehensive insights into the dynamics and broad-spectrum efficacy of neutralizing antibodies (NAbs) against emerging variants across different age groups, particularly in children, remain inadequate. Methods: Following the termination of China’s dynamic ‘zero-COVID-19’ policy in January 2023, which coincided with a widespread Omicron outbreak and numerous breakthrough infections, a longitudinal cohort study was established encompassing all age groups in Hubei, China. Follow-up assessments were conducted in March (Visit 1), June (Visit 2), and October (Visit 3) 2023. A total of 320 individuals were randomly selected and stratified into three age categories: children (<18 years, n = 80), adults (18–59 years, n = 167), and the elderly (≥60 years, n = 73). The NAbs against emerging SARS-CoV-2 variants BA.5, XBB.1.5, EG.5, and JN.1 were evaluated for each group. Trajectory modeling was employed to classify antibody trends into five categories: low-level stability, median-level stability, high-level stability, early increase, and late increase. Results: In March 2023, children exhibited significantly higher NAb levels against BA.5, XBB.1.5, EG.5, and JN.1 compared to adults and the elderly. However, these levels rapidly declined. From June to October 2023, no significant difference in NAb levels was observed between children and the other age groups. Regarding the broad-spectrum effectiveness of NAbs, the effectiveness in children was comparable to that of adults and the elderly in March 2023. However, from June to October 2023, children’s effectiveness became significantly lower than that of the other age groups. Trajectory analysis revealed that the highest proportions of high-level stability (31.3%) and median-level stability (42.5%) were observed among children. In contrast, adults and the elderly were most commonly categorized into the early increase (adult 46.7%, elderly 49.3%) and median-level stability (adult 22.1%, elderly 20.5%) categories. Conclusions: Although children initially demonstrate higher levels of NAbs, these levels decrease more rapidly than in adults and the elderly, eventually equalizing in later stages of recovery. Furthermore, the broad-spectrum effectiveness of NAbs in children is narrower than in other age groups. These findings suggest that children are at an elevated risk of infection with newly emerging variants, underscoring the urgent need to intensify focus on reinfections among children and develop tailored strategies to protect this vulnerable population. Full article

Soil salinization poses a severe threat to agricultural sustainability in the Yellow River Delta, where conventional spectral indices are limited by vegetation interference and seasonal dynamics in coastal saline-alkali landscapes. To address this, we developed an inversion framework integrating spectral indices and vegetation temporal features, combining multi-temporal Sentinel-2 optical data (January 2024–March 2025), Sentinel-1 SAR data, and terrain covariates. The framework employs Savitzky–Golay (SG) filtering to extract vegetation temporal indices—including NDVI temporal extremum and principal component features, capturing salt stress response mechanisms beyond single-temporal spectral indices. Based on 119 field samples and Variable Importance in Projection (VIP) feature selection, three ensemble models (XGBoost, CatBoost, LightGBM) were constructed under two strategies: single spectral features versus fused spectral and vegetation temporal features. The key results demonstrate the following: (1) The LightGBM model with fused features achieved optimal validation accuracy (R² = 0.77, RMSE = 0.26 g/kg), outperforming single-feature models by 13% in R². (2) SHAP analysis identified vegetation-related factors as key predictors, revealing a negative correlation between peak biomass and salinity accumulation, and the summer crop growth process affects soil salinization in the following spring. (3) The fused strategy reduced overestimation in low-salinity zones, enhanced model robustness, and significantly improved spatial gradient continuity. This study confirms that vegetation phenological features effectively mitigate agricultural interference (e.g., tillage-induced signal noise) and achieve high-resolution salinity mapping in areas where traditional spectral indices fail. The multi-temporal integration framework provides a replicable methodology for monitoring coastal salinization under complex land cover conditions. Full article

The object detection model for synthetic aperture radar (SAR) images needs to have strong generalization ability and more stable detection performance due to the complex scattering mechanism, high sensitivity of the orientation angle, and susceptibility to speckle noise. Visual large models possess strong generalization capabilities for natural image processing, but their application to SAR imagery remains relatively rare. This paper attempts to introduce a visual large model into the SAR object detection task, aiming to alleviate the problems of weak cross-domain generalization and poor adaptability to few-shot samples caused by the characteristics of SAR images in existing models. The proposed model comprises an image encoder, an attention module, and a detection decoder. The image encoder leverages the pre-trained Segment Anything Model (SAM) for effective feature extraction from SAR images. An Adaptive Channel Interactive Attention (ACIA) module is introduced to suppress SAR speckle noise. Further, a Dynamic Tandem Attention (DTA) mechanism is proposed in the decoder to integrate scale perception, spatial focusing, and task adaptation, while decoupling classification from detection for improved accuracy. Leveraging the strong representational and few-shot adaptation capabilities of large pre-trained models, this study evaluates their cross-domain and few-shot detection performance on SAR imagery. For cross-domain detection, the model was trained on AIR-SARShip-1.0 and tested on SSDD, achieving an mAP50 of 0.54. For few-shot detection on SAR-AIRcraft-1.0, using only 10% of the training samples, the model reached an mAP50 of 0.503. Full article

Inshore ship detection in synthetic aperture radar (SAR) imagery poses significant challenges due to the high density and diversity of ships. However, low inter-object backscatter contrast and blurred boundaries of docked ships often result in performance degradation for traditional object detection methods, especially under complex backgrounds and low signal-to-noise ratio (SNR) conditions. To address these issues, this paper proposes a novel detection framework, the Dynamic Weighted Joint Time–Frequency Feature Fusion DEtection TRansformer (DETR) Model (DWTF-DETR), specifically designed for SAR-based ship detection in inshore areas. The proposed model integrates a Dual-Domain Feature Fusion Module (DDFM) to extract and fuse features from both SAR images and their frequency-domain representations, enhancing sensitivity to both high- and low-frequency target features. Subsequently, a Dual-Path Attention Fusion Module (DPAFM) is introduced to dynamically weight and fuse shallow detail features with deep semantic representations. By leveraging an attention mechanism, the module adaptively adjusts the importance of different feature paths, thereby enhancing the model’s ability to perceive targets with ambiguous structural characteristics. Experiments conducted on a self-constructed inshore SAR ship detection dataset and the public HRSID dataset demonstrate that DWTF-DETR achieves superior performance compared to the baseline RT-DETR. Specifically, the proposed method improves mAP@50 by 1.60% and 0.72%, and F1-score by 0.58% and 1.40%, respectively. Moreover, comparative experiments show that the proposed approach outperforms several state-of-the-art SAR ship detection methods. The results confirm that DWTF-DETR is capable of achieving accurate and robust detection in diverse and complex maritime environments. Full article

Landslides constitute major geohazards that threaten human life, property, and ecological environments; it is imperative to acquire their location information accurately and in a timely manner. Interferometric Synthetic Aperture Radar (InSAR) has been demonstrated to be capable of acquiring subtle surface deformation with high precision and is widely applied to wide-area landslide detection. However, after obtaining InSAR deformation rates, visual interpretation is conventionally employed in landslide detection, which is characterized by significant temporal consumption and labor-intensive demands. Despite advancements that have been made through cluster analysis, hotspot analysis, and deep learning, persistent challenges such as low intelligence levels and weak generalization capabilities remain unresolved. In this study, we propose an improved Faster R-CNN model to achieve automatic detection of slow-moving landslides based on InSAR Line of Sight (LOS) annual rates in the upper and middle reaches of the Jinsha River Basin. The model incorporates a ResNet-34 backbone network, Feature Pyramid Network (FPN), and Convolutional Block Attention Module (CBAM) to effectively extract multi-scale features and enhance focus on subtle surface deformation regions. This model achieved test set performance metrics of 93.56% precision, 97.15% recall, and 93.6% F1-score. The proposed model demonstrates robust detection performance for slow-moving landslides, and through comparative analysis with the detection results of hotspot analysis and K-means clustering, it is verified that this method has strong generalization ability in the representative landslide-prone areas of the Qinghai–Tibet Plateau. This approach can support dynamic updates of regional slow-moving landslide inventories, providing crucial technical support for the detection of landslides. Full article

Understanding the atomistic basis of multi-layer mechanisms employed by broadly reactive neutralizing antibodies of the SARS-CoV-2 spike protein without directly blocking receptor engagement remains an important challenge in coronavirus immunology. Class 4 antibodies represent an intriguing case: they target a deeply conserved, cryptic epitope on the receptor-binding domain yet exhibit variable neutralization potency across subgroups F1 (CR3022, EY6A, COVA1-16), F2 (DH1047), and F3 (S2X259). The molecular basis for this variability is not fully understood. Here, we employed a multi-modal computational approach integrating atomistic and coarse-grained molecular dynamics simulations, binding free energy calculations, mutational scanning, and dynamic network analysis to elucidate how these antibodies engage the receptor-binding domain (RBD) of the SARS-CoV-2 spike protein and influence its function. Our results reveal that neutralization efficacy arises from the interplay of direct interfacial interactions and allosteric effects. Group F1 antibodies (CR3022, EY6A, COVA1-16) primarily operate via classic allostery, modulating flexibility in RBD loop regions to indirectly interfere with the ACE2 receptor binding through long-range effects. Group F2 antibody DH1047 represents an intermediate mechanism, combining partial steric hindrance—through engagement of ACE2-critical residues T376, R408, V503, and Y508—with significant allosteric influence, facilitated by localized communication pathways linking the epitope to the receptor interface. Group F3 antibody S2X259 achieves potent neutralization through a synergistic mechanism involving direct competition with ACE2 and localized allosteric stabilization, albeit with potentially increased escape vulnerability. Dynamic network analysis identified a conserved “allosteric ring” within the RBD core that serves as a structural scaffold for long-range signal propagation, with antibody-specific extensions modulating communication to the ACE2 interface. These findings support a model where Class 4 neutralization strategies evolve through the refinement of peripheral allosteric connections rather than epitope redesign. This study establishes a robust computational framework for understanding the atomistic basis of neutralization activity and immune escape for Class 4 antibodies, highlighting how the interplay of binding energetics, conformational dynamics, and allosteric modulation governs their effectiveness against SARS-CoV-2. Full article

This research provides valuable insights into the application of mathematical modeling to real-world scenarios, as exemplified by the COVID-19 pandemic. After data collection, the preparation stage included exploratory analysis, standardization and normalization, computation, and validation. A mathematical model initially developed for COVID-19 dynamics in Romania was subsequently applied to data from Italy and Switzerland during the same time interval. The model is structured as a multiple-input single-output (MISO) system, where the inputs underwent a neural network-based training stage to address inconsistencies in the acquired data. In parallel, an ARMAX model was employed to capture the stochastic nature of the epidemic process. Results demonstrate that the Romanian-based model generalized effectively across the three countries, achieving a strong predictive accuracy (forecast accuracy > 98.59%). Importantly, the model maintained robust performance despite significant cross-country differences in testing strategies, policy measures, timing of initial cases, and imported infections. This work contributes a novel perspective by showing that a unified data-driven modeling framework can be transferable across heterogeneous contexts. More broadly, it underscores the potential of integrating mathematical modeling with predictive analytics to support evidence-based decision-making and strengthen preparedness for future global health crises. Full article

Dynamic susceptibility assessment is essential for mitigating evolving landslide risks in alpine gorge regions. To address the static limitations and unit mismatch issues in conventional landslide susceptibility assessments in alpine gorge regions, this study proposes a dynamic framework integrating time-series InSAR-derived deformation. Applied to the Xinlong–Kangding section of the Yalong River, annual surface deformation velocities were retrieved using SBAS-InSAR with Sentinel-1 data, identifying 24 active landslide zones (>25 mm/a). The Geodetector model quantified the spatial influence of 18 conditioning factors, highlighting deformation velocity as the second most significant (q = 0.21), following soil type. Incorporating historical landslide data and InSAR deformation zones, slope unit delineation was optimized to construct a refined sample dataset. A Random Forest model was then used to assess the contribution of deformation factors. Results show that integrating InSAR data substantially improved model performance: “Very High” risk landslides increased from 67.21% to 87.01%, the AUC score improved from 0.9530 to 0.9798, and the Kappa coefficient increased from 0.7316 to 0.8870. These results demonstrate the value of InSAR-based dynamic monitoring in enhancing landslide susceptibility mapping, particularly for spatial clustering, classification precision, and model robustness. This approach offers a more efficient dynamic evaluation pathway for dynamic assessment and early warning of landslide hazards in mountainous regions. Full article

Synthetic aperture radar (SAR), with its all-weather and all-day active imaging capability, has become indispensable for geoscientific analysis and socio-economic applications. Despite advances in deep learning–based object detection, the rapid and accurate detection of small objects in SAR imagery remains a major challenge due to their extremely limited pixel representation, blurred boundaries in dense distributions, and the imbalance of positive–negative samples during training. Recently, vision–language models such as Contrastive Language-Image Pre-Training (CLIP) have attracted widespread research interest for their powerful cross-modal semantic modeling capabilities. Nevertheless, their potential to guide precise localization and detection of small objects in SAR imagery has not yet been fully exploited. To overcome these limitations, we propose the CLIP-Driven Adaptive Tiny Object Detection Diffusion Network (CDATOD-Diff). This framework introduces a CLIP image–text encoding-guided dynamic sampling strategy that leverages cross-modal semantic priors to alleviate the scarcity of effective positive samples. Furthermore, a generative diffusion-based module reformulates the sampling process through iterative denoising, enhancing contextual awareness. To address regression instability, we design a Balanced Corner–IoU (BC-IoU) loss, which decouples corner localization from scale variation and reduces sensitivity to minor positional errors, thereby stabilizing bounding box predictions. Extensive experiments conducted on multiple SAR and optical remote sensing datasets demonstrate that CDATOD-Diff achieves state-of-the-art performance, delivering significant improvements in detection robustness and localization accuracy under challenging small-object scenarios with complex backgrounds and dense distributions. Full article

Object detection is used to automatically identify and locate specific objects within images or videos for applications like autonomous driving, security surveillance, and medical imaging. Protecting object detection models against adversarial attacks, particularly malicious patches, is crucial to ensure reliable and safe performance in safety-critical applications, where misdetections can lead to severe consequences. Existing defenses against patch attacks are primarily designed for stationary scenes and struggle against adversarial image patches that vary in scale, position, and orientation in dynamic environments.In this paper, we introduce SAR, a patch-agnostic defense scheme based on image preprocessing that does not require additional model training. By integration of the patch-agnostic detection frontend with an additional broken pixel restoration backend, Segment and Recover (SAR) is developed for the large-mask-covered object-hiding attack. Our approach breaks the limitation of the patch scale, shape, and location, accurately localizes the adversarial patch on the frontend, and restores the broken pixel on the backend. Our evaluations of the clean performance demonstrate that SAR is compatible with a variety of pretrained object detectors. Moreover, SAR exhibits notable resilience improvements over state-of-the-art methods evaluated in this paper. Our comprehensive evaluation studies involve diverse patch types, such as localized-noise, printable, visible, and adaptive adversarial patches. Full article

Search and rescue (SAR) operations with unmanned aerial vehicles (UAVs) have been the subject of numerous scientific studies. Their effectiveness relies on intelligent and efficient path planning. Not only can they save expensive resources, they can minimize potential risks for the rescue team. This paper deals with optimal path planning for automated UAV-SAR operations, tailored specifically to the challenging inter-tidal environment of the Wadden Sea. The aim is to minimize the search time and maximize the discovery probability of lost persons (LPs) with intelligent UAV path-planning strategies. To achieve this, first a dynamic probability map (PM) of the lost person’s possible location is generated. Two distinct methods are evaluated for this purpose: Monte Carlo simulations (MCSs), and a more efficient Markov chain (MAC) model. The PM then directly informs the UAV’s decision-making process. Then, several automated search strategies are systematically evaluated and compared in a comprehensive simulation study, from simple coverage patterns to advanced PM-driven algorithms. MAC-generated PMs prove to provide a fast and reliable foundation for time-critical applications such as SAR operations. Additionally, PM-based search strategies outperform standard search patterns, especially in larger search regions. Furthermore, the evaluation is extended to multi-UAV scenarios, showing in this case that an area-segmentation approach is most effective. The results validate and provide a considerable contribution for an efficient, time-critical framework for UAV deployment in complex, real-world SAR operations. Full article