Search Results (200)

The integrity of subsea oil and gas pipelines is essential for offshore safety and environmental protection. Conventional leak detection approaches, such as manual inspection and indirect sensing, are often costly, time-consuming, and prone to subjectivity, motivating the development of automated methods. In this study, we present a deep learning-based framework for detecting underwater leaks using images acquired in controlled experiments designed to reproduce representative conditions of subsea monitoring. The dataset was generated by simulating both gas and liquid leaks in a water tank environment, under scenarios that mimic challenges observed during Remotely Operated Vehicle (ROV) inspections along the Brazilian coast. It was further complemented with artificially generated synthetic images (Stable Diffusion) and publicly available subsea imagery. Multiple Convolutional Neural Network (CNN) architectures, including VGG16, ResNet50, InceptionV3, DenseNet121, InceptionResNetV2, EfficientNetB0, and a lightweight custom CNN, were trained with transfer learning and evaluated on validation and blind test sets. The best-performing models achieved stable performance during training and validation, with macro F1-scores above 0.80, and demonstrated improved generalization compared to traditional baselines such as VGG16. In blind testing, InceptionV3 achieved the most balanced performance across the three classes when trained with synthetic data and augmentation. The study demonstrates the feasibility of applying CNNs for vision-based leak detection in complex underwater environments. A key contribution is the release of a novel experimentally generated dataset, which supports reproducibility and establishes a benchmark for advancing automated subsea inspection methods. Full article

Housing vacancy rate is a key indicator for evaluating urban sustainable development. Due to rapid urbanization, population outflow and insufficient industrial support, the housing vacancy problem is particularly prominent in China’s underdeveloped regions. However, the lack of official data and the limitations of traditional survey methods restrict in-depth research. This study proposes a vacancy rate estimation method based on Baidu Street View residential exterior images and deep learning technology. Taking Nanning, Guangxi as a case study, an automatic discrimination model for residential vacancy status is constructed by identifying visual clues such as window occlusion, balcony debris accumulation, and facade maintenance status. The study first uses Baidu Street View API to collect images of residential communities in Nanning. After manual annotation and field verification, a labeled dataset is constructed. A pre-trained deep learning model (ResNet50) is applied to estimate the vacancy rate of the community after fine-tuning with labeled street view images of Nanning’s residential communities. GIS spatial analysis is combined to reveal the spatial distribution pattern and influencing factors of the vacancy rate. The results show that street view images can effectively capture vacancy characteristics that are difficult to identify with traditional remote sensing and indirect indicators, providing a refined data source and method innovation for housing vacancy research in underdeveloped regions. The study further found that the residential vacancy rate in Nanning showed significant spatial differentiation, and the vacancy driving mechanism in the old urban area and the emerging area was significantly different. This study expands the application boundaries of computer vision in urban research and fills the research gap on vacancy issues in underdeveloped areas. Its results can provide a scientific basis for the government to optimize housing planning, developers to make rational investments, and residents to make housing purchase decisions, thus helping to improve urban sustainable development and governance capabilities. Full article

Objectives: Paratubal cysts (PTCs) are embryological remnants and are potentially hormonally responsive. Since hyperandrogenism (HA) is representative of polycystic ovary syndrome (PCOS), we examined whether biochemical hyperandrogenism is associated with PTCs in women with PCOS and if body mass index (BMI) and insulin resistance (IR) mediate this association. Methods: This retrospective study included 577 women diagnosed with PCOS at a tertiary academic center from 2010 to 2018. Clinical data included age at diagnosis, BMI, and diagnoses of hypertension, non-alcoholic fatty liver disease, and metabolic syndrome. Laboratory measures included total testosterone, sex hormone-binding globulin, anti-Müllerian hormone, luteinizing hormone, fasting glucose, insulin, and triglycerides (TG). Derived indices included a free androgen index (FAI), homeostasis model assessment of insulin resistance (HOMA-IR), and fasting glucose-to-insulin ratio. PTCs were identified through imaging or surgical findings. Structural equation modeling (SEM) assessed direct and indirect relationships between FAI, BMI, HOMA-IR, and PTCs, while adjusting for diagnostic age. Results: PTCs were identified in 2.77% of participants. BMI, FAI, TG, and IR indices were significantly higher for women with PTCs than those without PTCs. SEM revealed significant indirect effects of FAI on PTCs via BMI and HOMA-IR. The direct effect was negative, resulting in a non-significant total effect. A sensitivity model using HOMA-IR as the predictor showed a significant direct effect on PTCs without mediation via FAI. Conclusions: Biochemical HA may influence PTC development in PCOS through metabolic pathways, establishing the need to consider metabolic context when evaluating adnexal cysts in hyperandrogenic women. Full article

Temperature is a key physiological indicator of plant health, influenced by factors including water status, disease and developmental stage. Monitoring changes in multiple factors is helpful for early diagnosis of plant growth. However, there are a variety of complex light interference phenomena in the greenhouse, so traditional detection methods cannot meet effective online monitoring of strawberry health status without manual intervention. Therefore, this paper proposes a leaf soft-sensing method based on a thermal infrared imaging sensor and adaptive image screening Internet of Things system, with additional sensors to realize indirect and rapid monitoring of the health status of a large range of strawberries. Firstly, a fuzzy comprehensive evaluation model is established by analyzing the environmental interference terms from the other sensors. Secondly, through the relationship between plant physiological metabolism and canopy temperature, a growth model is established to predict the growth period of strawberries based on canopy temperature. Finally, by deploying environmental sensors and solar height sensors, the image acquisition node is activated when the environmental interference is less than the specified value and the acquisition is completed. The results showed that the accuracy of this multiple sensors system was 86.9%, which is 30% higher than the traditional model and 4.28% higher than the latest advanced model. It makes it possible to quickly and accurately assess the health status of plants by a single factor without in-person manual intervention, and provides an important indication of the early, undetectable state of strawberry disease, based on remote operation. Full article

Alzheimer’s disease (AD) involves the accumulation of amyloid-β (Aβ) plaques, whose quantification plays a central role in understanding disease progression. Automated segmentation of Aβ deposits in histopathological micrographs enables large-scale analyses but is hindered by the high cost of detailed pixel-level annotations. Weakly supervised learning offers a promising alternative by leveraging coarse or indirect labels to reduce the annotation burden. We evaluated a weakly supervised approach to segment and analyze thioflavin-S-positive parenchymal amyloid pathology in AD and age-matched brains. Our pipeline integrates three key components, each designed to operate under weak supervision. First, robust preprocessing (including retrospective multi-image illumination correction and gradient-based background estimation) was applied to enhance image fidelity and support training, as models rely more on image features. Second, class activation maps (CAMs), generated by a compact deep classifier SqueezeNet, were used to identify, and coarsely localize amyloid-rich parenchymal regions from patch-wise image labels, serving as spatial priors for subsequent refinement without requiring dense pixel-level annotations. Third, a patch-based convolutional neural network, U-Net, was trained on synthetic data generated from micrographs based on CAM-derived pseudo-labels via an extensive object-level augmentation strategy, enabling refined whole-image semantic segmentation and generalization across diverse spatial configurations. To ensure robustness and unbiased evaluation, we assessed the segmentation performance of the entire framework using patient-wise group k-fold cross-validation, explicitly modeling generalization across unseen individuals, critical in clinical scenarios. Despite relying on weak labels, the integrated pipeline achieved strong segmentation performance with an average Dice similarity coefficient (≈0.763) and Jaccard index (≈0.639), widely accepted metrics for assessing segmentation quality in medical image analysis. The resulting segmentations were also visually coherent, demonstrating that weakly supervised segmentation is a viable alternative in histopathology, where acquiring dense annotations is prohibitively labor-intensive and time-consuming. Subsequent morphometric analyses on automatically segmented Aβ deposits revealed size-, structural complexity-, and global geometry-related differences across brain regions and cognitive status. These findings confirm that deposit architecture exhibits region-specific patterns and reflects underlying neurodegenerative processes, thereby highlighting the biological relevance and practical applicability of the proposed image-processing pipeline for morphometric analysis. Full article

The goal of on-device continual learning is to enable models to adapt to streaming data without forgetting previously acquired knowledge, even with limited computational resources and memory constraints. Recent research has demonstrated that weighted regularization-based methods are constrained by indirect knowledge preservation and sensitive hyperparameter settings, and dynamic architecture methods are ill-suited for on-device environments due to increased resource consumption as the structure scales. In order to compensate for these limitations, replay-based continuous learning, which maintains a compact structure and stable performance, is gaining attention. The limitations of replay-based continuous learning are (1) the limited amount of historical training data that can be stored due to limited memory capacity, and (2) the computational resources of on-device systems are significantly lower than those of servers or cloud infrastructures. Consequently, designing strategies that balance the preservation of past knowledge with rapid and cost-effective updates of model parameters has become a critical consideration in on-device continual learning. This paper presents an empirical survey of replay-based continual learning studies, considering the nearest class mean classifier with replay-based sparse weight updates as a representative method for validating the feasibility of diverse edge devices. Our empirical comparison of standard benchmarks, including CIFAR-10, CIFAR-100, and TinyImageNet, deployed on devices such as Jetson Nano and Raspberry Pi, showed that the proposed representative method achieved reasonable accuracy under limited buffer sizes compared with existing replay-based techniques. A significant reduction in training time and resource consumption was observed, thereby supporting the feasibility of replay-based on-device continual learning in practice. Full article

Seagrass constitutes a vital component of coastal ecosystems, providing a wide array of ecosystem services. The accurate measurement of the seagrass frontal area is crucial for assessing its capacity to inhibit water flow and reduce wave energy; however, few effective indirect methods exist. To address this limitation, we developed an indirect method that combines the Mask R-CNN model with a dual-reference approach for detecting seagrass and estimating its frontal area. A laboratory-scale underwater camera experiment generated an experimental dataset, which was partitioned into training, validation, and test sets. Following training, evaluation metrics—including IoU, accuracy, precision, recall, and F1-score—approached their upper limits and remained within acceptable ranges. Validation on real seagrass images confirmed satisfactory performance, albeit with slightly lower metrics than those observed in the experimental dataset. Furthermore, the method estimated seagrass frontal areas with errors below 10% (maximum 7.68% and minimum –0.43%), thereby demonstrating high accuracy by accounting for seagrass bending under flowing water conditions. Additionally, we showed that the indirect measurement significantly influences estimations of the seagrass bending height and wave height reduction capacity, mitigating the overestimation associated with traditional direct methods. Thus, this indirect approach offers a promising, environmentally friendly alternative that overcomes the limitations of conventional techniques. Full article

Soil texture, defined by the proportions of sand, silt, and clay particles in the soil, is one of the most essential physical properties of soil. High-resolution soil texture data can provide critical parameter support for soil hydrological modeling, agricultural production management, and ecosystem assessment. In digital soil mapping, previous studies often predicted the sand, silt, and clay contents in soil and then indirectly calculated soil texture. Currently, approaches that directly map soil texture by classification modeling are gaining increasing attention due to the decreased error from data conversion, but few studies have systematically compared these two methods yet. In this study, we comprehensively assessed the performance of direct and indirect predicting soil texture using four machine learning algorithms (e.g., extreme gradient boosting, random forest, gradient boosting decision tree, and extremely randomized tree) with 190 covariates from the Digital Elevation Model, Sentinel-1/2 satellite images, and classification maps and generated a 10 m resolution soil texture map based on 405 topsoil (0–20 cm) sample data collected in Suichuan County, China. The results showed that compared with indirect predictions, direct predictions improved overall accuracy (OA) by 20.57–44.19% and the Kappa coefficient (Kappa) by 0.220–0.402. Among the models used, the XGB model achieved the highest accuracy (OA: 0.948; Kappa: 0.931) and the lowest uncertainty (confusion index: 0.052). The direct prediction map (nine classes recorded) exhibited more detailed and diverse spatial distribution patterns than the indirect prediction map (six classes recorded), aligning better with the actual environment. Based on accuracy validation and spatial distribution, the performance of the XGB model was best during direct prediction. The Shapley additive explanation from the XGB model revealed that the normalized height and stream power indices were the most significant factors driving the soil texture in the study area. Our results provide a reference for future studies on soil texture mapping using machine learning models. Full article

Accurate imaging methods are important for understanding electrodeposition phenomena in metal batteries. Among the suitable imaging methods for this task is magnetic resonance imaging (MRI), which is a very powerful radiological diagnostic method. In this study, MR microscopy was used to image electroplating in a lithium symmetric cell, which was used as a model for a lithium-metal battery. Lithium electrodeposition in this cell was studied by sequential 3D ¹H MRI of 1 M LiPF₆ in EC/DMC electrolyte under different charging conditions, which resulted in different dynamics of the amount of electroplated lithium and its structure. The acquired images depicted the electrolyte distribution, so that the images of deposited lithium that did not give a detectable signal corresponded to the negatives of these images. With this indirect MRI, phenomena such as the transition from a mossy to a dendritic structure at Sand’s time, the growth of whiskers, the growth of dendrites with arborescent structure, the formation of dead lithium, and the formation of gas due to electrolyte decomposition were observed. In addition, the effect of charge and discharge cycles on electrodeposition was also studied. It was found that it is difficult to correctly predict the occurrence of these phenomena based on charging conditions alone, as seemingly identical conditions resulted in different results. Full article

The study investigates the influence of an air-entraining admixture on the properties and pore structure of ordinary concrete. The aim was to examine how modifications to the concrete mix affect compressive strength, ultrasonic pulse velocity, and resistance to freeze–thaw cycles. Concrete samples with varying admixture dosages (0.00–1.50% of cement mass) were tested for mechanical properties and pore structure. Freeze–thaw resistance was assessed using both direct (PN-B-06265) and indirect methods (EN 480-11), while pore characteristics were evaluated via computer-aided image analysis. Results show that increasing the admixture dosage enhances freeze–thaw resistance by refining the pore structure—particularly by increasing the content of micropores below 0.3 mm—while simultaneously reducing compressive strength and ultrasonic velocity. Statistical analysis revealed that pore parameters such as total air content, specific surface area, and spacing factor significantly correlate with concrete performance. The regression models confirmed that compressive strength and ultrasonic velocity are negatively impacted by increased pore volume, while freeze–thaw resistance improves due to a more favorable pore size distribution. The findings demonstrate that optimizing the admixture dosage can effectively balance durability and mechanical performance, and that quantitative stereological parameters provide a valuable basis for predicting the behavior of air-entrained concrete. Full article

Oriented ship detection in Synthetic Aperture Radar (SAR) images has broad applications in maritime surveillance and other fields. While deep learning advancements have significantly improved ship detection performance, persistent challenges remain for existing methods. These include the inherent misalignment between regression and classification tasks and the boundary discontinuity problem in oriented object detection. These issues hinder efficient and accurate ship detection in complex scenarios. To address these challenges, we propose TIAR-SAR, a novel oriented SAR ship detector featuring a task interaction head and composite angle regression. First, we propose a task interaction detection head (Tihead) capable of predicting both oriented bounding boxes (OBBs) and horizontal bounding boxes (HBBs) simultaneously. Within the Tihead, a “decompose-then-interact” structure is designed. This structure not only mitigates feature misalignment but also promotes feature interaction between regression and classification tasks, thereby enhancing prediction consistency. Second, we propose a joint angle refinement mechanism (JARM). The JARM addresses the non-differentiability problem of the traditional rotated Intersection over Union (IoU) loss through the design of a composite angle regression loss (CARL) function, which strategically combines direct and indirect angle regression methods. A boundary angle correction mechanism (BACM) is then designed to enhance angle estimation accuracy. During inference, BACM dynamically replaces an object’s OBB prediction with its corresponding HBB if the OBB exhibits excessive angle deviation when the angle of the object is near the predefined boundary. Finally, the performance and applicability of the proposed methods are evaluated through extensive experiments on multiple public datasets, including SRSDD, HRSID, and DOTAv1. Experimental results derived from the use of the SRSDD dataset demonstrate that the mAP50 of the proposed method reaches 63.91%, an improvement of 4.17% compared with baseline methods. The detector achieves 17.42 FPS on 1024 × 1024 images using an RTX 2080 Ti GPU, with a model size of only 21.92 MB. Comparative experiments with other state-of-the-art methods on the HRSID dataset demonstrate the proposed method’s superior detection performance in complex nearshore scenarios. Furthermore, when further tested on the DOTAv1 dataset, the mAP50 can reach 79.1%. Full article

A composite source model consists of an intrinsic state and an extrinsic observation. The fundamental performance limit of reproducing the intrinsic state is characterized by the indirect rate–distortion function. In a remote classification application, a source encoder encodes the extrinsic observation (e.g., image) into bits, and a source decoder plays the role of a classifier that reproduces the intrinsic state (e.g., label of image). In this work, we characterize the general structure of the optimal transition probability distribution, achieving the indirect rate–distortion function. This optimal solution can be interpreted as a “soft classifier”, which generalizes the conventionally adopted “classify-then-compress” scheme. We then apply the soft classification to aid the lossy compression of the extrinsic observation of a composite source. This leads to a coding scheme that exploits the soft classifier to guide reproduction, outperforming existing coding schemes without classification or with hard classification. Full article

Biomass burning (BB) aerosols are the largest source of absorbing aerosols on Earth. Coupled with marine stratocumulus clouds (MSC), their radiative effects are enhanced and can cause cloud property changes (first indirect effect) or cloud burn-off and warm up the atmospheric column (semi-direct effect). Nevertheless, the derivation of their quantity and optical properties in the presence of MSC clouds is confounded by the uncertainties in the retrieval of the underlying cloud properties. Therefore, a robust methodology is needed for the coupled retrievals of absorbing aerosol above clouds. Here, we present a new retrieval approach implemented for a Spectro radiometric multi-angle polarimetric airborne platform, the research scanning polarimeter (RSP), during the ORACLES campaign over the Southeast Atlantic Ocean. Our approach transforms the 1D measurements over multiple angles and wavelengths into a 3D image-like input, which is then processed using various deep learning (DL) schemes to yield aerosol single scattering albedos (SSAs), aerosol optical depths (AODs), aerosol effective radii, and aerosol complex refractive indices, together with cloud optical depths (CODs), cloud effective radii and variances. We present a comparison between the different DL approaches, as well as their comparison to existing algorithms. We discover that the Vision Transformer (ViT) scheme, traditionally used by natural language models, is superior to the ResNet convolutional Neural-Network (CNN) approach. We show good validation statistics on synthetic and real airborne data and discuss paths forward for making this approach flexible and readily applicable over multiple platforms. Full article

Photonic integrated interferometric imaging systems (PIISs) provide a compact solution for high-resolution Earth observation missions with stringent size, weight, and power (SWaP) constraints. As an indirect imaging method, a PIIS exhibits fundamentally different noise response characteristics compared to conventional remote sensing systems, and its imaging performance under practical operational scenarios has not been thoroughly investigated. The primary objective of this paper is to evaluate the operational capabilities of PIISs under remote sensing conditions. We (1) establish a signal-to-noise-ratio model for PIISs with balanced four-quadrature detection, (2) analyze the impacts of intensity noise and turbulent phase noise based on radiative transfer and turbulence models, and (3) simulate imaging performance with WorldView-3-like parameters. The results of the visibility signal-to-noise ratio (SNR) analysis demonstrate that the system’s minimum detectable fringe visibility is inversely proportional to the reciprocal of the sub-aperture intensity signal-to-noise ratio. When the integration time reaches 100 ms, the minimum detectable fringe visibility ranges between ${10}^{-2}$ and ${10}^{-3}$ (at 10 dB system efficiency). Imaging simulations demonstrate that recognizable image reconstruction requires integration times exceeding 10 ms for 10 cm baselines, achieving approximately 25 dB PSNR and 0.8 SSIM at 100 ms integration duration. These results may provide references for potential applications of photonic integrated interferometric imaging systems in remote sensing. Full article

Body awareness consists of aesthetic body image, functional body image, and interoception. Previous studies indicated a link between these components of body awareness and mental health. This study aims to clarify the relationship among these variables during the period of pubertal body changes. As puberty progresses, individuals’ perceptions of their bodies shift, which has been associated with a decline in mental health, according to the existing literature. To investigate this issue, a sample of 294 post-pubertal adolescents assigned female at birth completed assessments related to body awareness, mental health, psychosomatic symptoms, gender congruence, and eating disorders. A network analysis was conducted to illustrate the intricate interactions among the observed variables, and a mediation model was utilised to explore how body image influences overall health, with interoception and functional body image acting as mediators. The study identifies three key variables—body image, mental health, and interoception—as central within the network. Additionally, functional body image was significantly associated with other variables in the study. Ultimately, both direct and indirect effects of body image on mental health were found, mediated through interoception and functional body image. The clinical implications emphasise the importance of enhancing awareness of bodily sensations and functions to support psychological well-being, particularly during a developmental stage characterised by challenges related to body image due to rapid changes in puberty. Full article