Search Results (408)

Visual inspection of civil infrastructure for structural health assessment, as performed by structural engineers, is expensive and time-consuming. Therefore, automating this process is highly attractive, which has received significant attention in recent years. With the increasing capabilities of computers, deep neural networks have become a standard tool and can be used for structural health inspections. A key challenge, however, is the availability of reliable datasets. In this work, the U-net and DeepLab v3+ convolutional neural networks are trained on a synthetic Tokaido dataset. This dataset comprises images representative of data acquired by unmanned aerial vehicle (UAV) imagery and corresponding ground truth data. The data includes semantic segmentation masks for both categorizing structural elements (slabs, beams, and columns) and assessing structural damage (concrete spalling or exposed rebars). Data augmentation, including both image quality degradation (e.g., brightness modification, added noise) and image transformations (e.g., image flipping), is applied to the synthetic dataset. The selected neural network architectures achieve excellent performance, reaching values of 97% for accuracy and 87% for Mean Intersection over Union (mIoU) on the validation data. It also demonstrates promising results in the semantic segmentation of real-world structures captured in photographs, despite being trained solely on synthetic data. Additionally, based on the obtained results of semantic segmentation, it can be concluded that DeepLabV3+ outperforms U-net in structural component identification. However, this is not the case in the damage identification task. Full article

Non-destructive evaluation of reinforced concrete structures is essential for effective maintenance and safety assessments. This study explores the combined use of active microwave thermography and deep learning to detect and localize steel reinforcement within concrete elements. Numerical simulations were developed to model the thermal response of reinforced concrete subjected to microwave excitation, generating synthetic thermal images representing the surface temperature patterns of reinforced concrete, influenced by subsurface steel reinforcement. These images served as training data for a deep neural network designed to identify and localize rebar positions based on thermal patterns. The model was trained exclusively on simulation data and subsequently validated using experimental measurements obtained from large-format concrete slabs incorporating a structured layout of embedded steel reinforcement bars. Surface temperature distributions obtained through infrared imaging were compared with model predictions to evaluate detection accuracy. The results demonstrate that the proposed method can successfully identify the presence and approximate location of internal reinforcement without damaging the concrete surface. This approach introduces a new pathway for contactless, automated inspection using a combination of physical modeling and data-driven analysis. While the current work focuses on rebar detection and localization, the methodology lays the foundation for broader applications in non-destructive testing of concrete infrastructure. Full article

Street networks provide an invaluable source of information about the different temporal and spatial patterns emerging in our cities. These streets are often represented as graphs where intersections are modeled as nodes and streets as edges between them. Previous work has shown that raster representations of the original data can be created through a learning algorithm on low-dimensional representations of the street networks. In contrast, models that capture high-level urban network metrics can be trained through convolutional neural networks. However, the detailed topological data is lost through the rasterization of the street network, and the models cannot recover this information from the image alone, failing to capture complex street network features. This paper proposes a model capable of inferring good representations directly from the street network. Specifically, we use a variational autoencoder with graph convolutional layers and a decoder that generates a probabilistic, fully connected graph to learn latent representations that encode both local network structure and the spatial distribution of nodes. We train the model on thousands of street network segments and use the learned representations to generate synthetic street configurations. Finally, we proposed a possible application to classify the urban morphology of different network segments, investigating their common characteristics in the learned space. Full article

High-resolution ISAR (Inverse Synthetic Aperture Radar) imaging plays a crucial role in dynamic target monitoring for aerospace, maritime, and ground surveillance. Among various remote sensing techniques, ISAR is distinguished by its ability to produce high-resolution images of non-cooperative maneuvering targets. To meet the increasing demands for resolution and robustness, modern ISAR systems are evolving toward wideband and multi-channel architectures. In particular, multi-channel configurations based on large-scale receiving arrays have gained significant attention. In such systems, each receiving element functions as an independent spatial channel, acquiring observations from distinct perspectives. These multi-angle measurements enrich the available echo information and enhance the robustness of target imaging. However, this setup also brings significant challenges, including inter-channel coupling, high-dimensional joint signal modeling, and non-Gaussian, mixed-mode interference, which often degrade image quality and hinder reconstruction performance. To address these issues, this paper proposes a Hybrid Variational Bayesian Multi-Interference (HVB-MI) imaging algorithm based on a hierarchical Bayesian framework. The method jointly models temporal correlations and inter-channel structure, introducing a coupled processing strategy to reduce dimensionality and computational complexity. To handle complex noise environments, a Gaussian mixture model (GMM) is used to represent nonstationary mixed noise. A variational Bayesian inference (VBI) approach is developed for efficient parameter estimation and robust image recovery. Experimental results on both simulated and real-measured data demonstrate that the proposed method achieves significantly improved image resolution and noise robustness compared with existing approaches, particularly under conditions of sparse sampling or strong interference. Quantitative evaluation further shows that under the continuous sparse mode with a 75% sampling rate, the proposed method achieves a significantly higher Laplacian Variance (LV), outperforming PCSBL and CPESBL by 61.7% and 28.9%, respectively and thereby demonstrating its superior ability to preserve fine image details. Full article

There is a common need for optical sequence images with high spatiotemporal resolution. As a solution, Synthetic Aperture Radar (SAR)-to-optical translation tends to bring high temporal continuity of optical images and low interpretation difficulty of SAR images. Existing studies have been focused on converting a single SAR image into a single optical image, failing to utilize the advantages of repeated observations from SAR satellites. To make full use of periodic SAR images, it is proposed to investigate the sequential SAR-to-optical translation, which represents the first effort in this topic. To achieve this, a model based on a diffusion framework has been constructed, with twelve Transformer blocks utilized to effectively capture spatial and temporal features alternatively. A variational autoencoder is employed to encode and decode images, enabling the diffusion model to learn the distribution of features within optical image sequences. A conditional branch is specifically designed for SAR sequences to facilitate feature extraction. Additionally, the capture time is encoded and embedded into the Transformers. Two sequence datasets for the sequence translation task were created, comprising Sentinel-1 Ground Range Detected data and Sentinel-2 red/green/blue data. Our method was tested on new datasets and compared with three state-of-the-art single translation methods. Quantitative and qualitative comparisons validate the effectiveness of the proposed method in maintaining radiometric and spectral consistency. Full article

With the rapid advancement of deepfake technology, the detection of low-quality synthetic facial images has become increasingly challenging, particularly in cases involving low resolution, blurriness, or noise. Traditional detection methods often exhibit limited performance under such conditions. To address these limitations, this paper proposes a novel algorithm, YOLOv9-ARC, which is designed to enhance the accuracy of detecting low-quality fake facial images. The proposed algorithm introduces an innovative convolution module, Adaptive Kernel Convolution (AKConv), which dynamically adjusts kernel sizes to effectively extract image features, thereby mitigating the challenges posed by low resolution, blurriness, and noise. Furthermore, a hybrid attention mechanism, Convolutional Block Attention Module (CBAM), is integrated to amplify salient features while suppressing irrelevant information. Extensive experiments demonstrate that YOLOv9-ARC achieves a mean average precision (mAP) of 75.1% on the DFDC (DeepFake Detection Challenge) dataset, representing a 3.5% improvement over the baseline model. The proposed YOLOv9-ARC not only addresses the challenges of low-quality deepfake detection but also demonstrates significant improvements in accuracy within this domain. Full article

Chronobiotics represent a pharmacologically diverse group of substances, encompassing both experimental compounds and those utilized in clinical practice, which possess the capacity to modulate the parameters of circadian rhythms. These substances influence fluctuations in various physiological and biochemical processes, including the expression of core “clock” genes in model organisms and cell cultures, as well as the expression of clock-controlled genes. Despite their chemical heterogeneity, chronobiotics share the common ability to alter circadian dynamics. The concept of chronobiotic drugs has been recognized for over five decades, dating back to the discovery and detailed clinical characterization of the hormone melatonin. However, the field remains fragmented, lacking a unified classification system for these pharmacological agents. The current categorizations include natural chrononutrients, synthetic targeted circadian rhythm modulators, hypnotics, and chronobiotic hormones, yet no comprehensive repository of knowledge on chronobiotics exists. Addressing this gap, the development of the world’s first curated and continuously updated database of chronobiotic drugs—circadian rhythm modulators—accessible via the global Internet, represents a critical and timely objective for the fields of chronobiology, chronomedicine, and pharmacoinformatics/bioinformatics. The primary objective of this study is to construct a relational database, ChronobioticsDB, utilizing the Django framework and PostGreSQL as the database management system. The database will be accessible through a dedicated web interface and will be filled in with data on chronobiotics extracted and manually annotated from PubMed, Google Scholar, Scopus, and Web of Science articles. Each entry in the database will comprise a detailed compound card, featuring links to primary data sources, a molecular structure image, the compound’s chemical formula in machine-readable SMILES format, and its name according to IUPAC nomenclature. To enhance the depth and accuracy of the information, the database will be synchronized with external repositories such as ChemSpider, DrugBank, Chembl, ChEBI, Engage, UniProt, and PubChem. This integration will ensure the inclusion of up-to-date and comprehensive data on each chronobiotic. Furthermore, the biological and pharmacological relevance of the database will be augmented through synchronization with additional resources, including the FDA. In cases of overlapping data, compound cards will highlight the unique properties of each chronobiotic, thereby providing a robust and multifaceted resource for researchers and practitioners in the field. Full article

This paper addresses the challenges of sample scarcity and class imbalance in remote sensing image semantic segmentation by proposing a decoupled synthetic sample generation framework based on a latent diffusion model. The method consists of two stages. In the label generation stage, we fine-tune a pretrained latent diffusion model with LoRA to generate semantic label masks from textual descriptions. A novel proportion-aware loss function explicitly penalizes deviations from the desired class distribution in the generated mask. In the image generation stage, we use ControlNet to train a multi-condition image generation network that takes the synthesized mask, along with its text description, as input and produces a realistic remote sensing image. The base Stable Diffusion model’s weights remain frozen during this process, with the trainable ControlNet ensuring that outputs are structurally and semantically aligned with the input labels. This two-stage approach yields coherent image–mask pairs that are well-suited for training segmentation models. Experiments show that models trained on the synthetic samples produced by the proposed method achieve high visual quality and semantic consistency. The proportion-aware loss effectively mitigates the impact of minority classes, boosting segmentation performance on under-represented categories. Results also reveal that adding a suitable proportion of synthetic sample improves segmentation accuracy, whereas an excessive share can cause over-fitting or misclassification. Comparative tests across multiple models confirm the generality and robustness of the approach. Full article

Histopathological images represent a valuable data source for pathologists, who can provide clinicians with essential landmarks for complex pathologies. The development of sophisticated computational models for histopathological images has received significant attention in recent years, but most of them rely on free datasets. Materials and Methods: Motivated by this drawback, the authors created an original histopathological image dataset that resulted from an animal experimental model, acquiring images from normal female rats/rats with experimentally induced diabetes mellitus (DM)/rats who received an antidiabetic therapy with a synthetic compound (AD_SC). Images were acquired from vaginal, uterine, and ovarian samples from both MD and AD_DC specimens. The experiment received the approval of the Medical Ethics Committee of the “Gr. T. Popa” University of Medicine and Pharmacy, Iași, Romania (Approval No. 169/22.03.2022). The novelty of the study consists of the following aspects. The first is the use of a diabetes-induced animal model to evaluate the impact of an antidiabetic therapy with a synthetic compound in female rats, focusing on three distinct organs of the reproductive system (vagina, ovary, and uterus), to provide a more comprehensive understanding of how diabetes affects female reproductive health as a whole. The second comprises image classification with a custom-built convolutional neural network (CB-CNN), the extraction of textural features (contrast, entropy, energy, and homogeneity), and their classification with PyCaret Auto Machine Learning (AutoML). Results: Experimental findings indicate that uterine tissue, both for MD and AD_DC, can be diagnosed with an accuracy of 94.5% and 85.8%, respectively. The Linear Discriminant Analysis (LDA) classifier features indicate a high accuracy of 86.3% when supplied with features extracted from vaginal tissue. Conclusions: Our research underscores the efficacy of classifying with two AI algorithms, CNN and machine learning. Full article

Aza-crown ether structures have been proven to be effective in constructing fluorescent biosensors for selectively detecting and imaging alkali metal ions in biological environments. However, choosing the right aza-crown ether for a specific alkali metal ion remains challenging for synthetic chemists because theoretical guidance on the chelating activities between aza-crown ethers and alkali metal ions has not been available up to now. Predicting the physical properties of the chelator–metal complexations poses a greater challenge due to the numerous quantum mechanical functionals and basis sets to be used in any theoretical investigation. In this study, we report a theoretical investigation of different aza-crown ether structures and their selectivities to alkali metal ions via a novel relationship between the binding energy and charge transfer calculated using twelve different quantum mechanical methods, using a myriad of bases, within the Jacob’s Ladder of Chemical Accuracies. Furthermore, this report represents a guide for the synthetic chemist in the selection of aza-crown ethers in the capturing of specific alkali metal ions, primary objectives, while benchmarking different quantum mechanical calculations, as a secondary objective. Full article

Oil spills represent a serious threat to marine ecosystems. Remote sensing monitoring, especially based on synthetic aperture radar (SAR), have been extensively employed in marine environments due to its unique advantages. However, SAR images of marine oil spills exhibit challenges of weak boundaries, confusion with look-alike phenomena, and the difficulty of detecting small-scale targets. To address these issues, we propose LRA-UNet, a Lightweight Residual Attention UNet for semantic segmentation in SAR images. Our model integrates depthwise separable convolutions to reduce feature redundancy and computational cost, while adopting a residual encoder enhanced with the Simple Attention Module (SimAM) to improve the precise extraction of target features. Additionally, we design a joint loss function that incorporates Sobel-based edge information, emphasizing boundary features during training to enhance edge sharpness. Experimental results show that LRA-UNet achieves superior segmentation results, with a mIoU of 67.36%, surpassing the original UNet by 4.41%, and a 5.17% improvement in IoU for the oil spill category. These results confirm the effectiveness of our approach in accurately extracting oil spill regions from complex SAR imagery. Full article

Central Asia, and in particular Tajikistan, is one of the most geologically hazardous areas in the world, particularly in terms of seismicity, floods, and landslides. The majority of landslides that occur in the region are seismically induced. A notable site is the Baipaza landslide, which has been subject to deformation since the 1960s, with the most recent collapse occurring in 2002. The potential collapse of the landslide represents a significant risk to the nearby Baipaza hydroelectric dam, situated 5 km away, and has the potential to create widespread challenges for the entire region. The objective of this work is to provide a comprehensive characterization of the Baipaza landslide through the utilization of satellite remote-sensing techniques, exploiting both Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-2 optical images freely available from the European Space Agency’s (ESA) Copernicus project. The employment of these two techniques enables the acquisition of insights into the distinctive characteristics and dynamics of the landslide, including the displacement rates up to 246 mm/year in the horizontal component; the precise mapping of landslide boundaries and the identification of distinct sectors with varying deformation patterns; and an estimation of the volume involved within the landslide, which is approximately of 1 billion m³. Full article

Deep neural networks (DNNs) have shown strong performance in synthetic aperture radar (SAR) image classification. However, their “black-box” nature limits interpretability and poses challenges for robustness, which is critical for sensitive applications such as disaster assessment, environmental monitoring, and agricultural insurance. This study systematically evaluates the adversarial robustness of five representative DNNs (VGG11/16, ResNet18/101, and A-ConvNet) under a variety of attack and defense settings. Using eXplainable AI (XAI) techniques and attribution-based visualizations, we analyze how adversarial perturbations and adversarial training affect model behavior and decision logic. Our results reveal significant robustness differences across architectures, highlight interpretability limitations, and suggest practical guidelines for building more robust SAR classification systems. We also discuss challenges associated with large-scale, multi-class land use and land cover (LULC) classification under adversarial conditions. Full article

Citrus sooty mould caused by Capnodium spp. alters the quality of fruits on the tree and affects their productivity. Past laboratory and hand-held spectrometry tests have concluded that sooty mould exhibits a typical spectral response in the near-infrared spectrum region. For this reason, this study aims at developing an automatic method for remote sensing of this disease, combining 10 m spatial resolution Sentinel-2 satellite images and 0.25 m spatial resolution orthophotos to identify sooty mould infestation levels in small orchards, common in Mediterranean conditions. Citrus orchards of the Comunitat Valenciana region (Spain) underwent field inspection in 2022 during two months of minimum (August) and maximum (October) infestation. The inspectors categorised their observations according to three levels of infestation in three representative positions of each orchard. Two synthetic images condensing the monthly information were generated for both periods. A filtering algorithm was created, based on high-resolution images, to select informative pixels in the lower resolution images. The data were used to evaluate the performance of a Random Forest classifier in predicting intensity levels through cross-validation. Combining the information from medium- and high-resolution images improved the overall accuracy from 0.75 to 0.80, with mean producer’s accuracies of above 0.65 and mean user’s accuracies of above 0.78. Bowley–Yule skewness coefficients were +0.50 for the overall accuracy and +0.28 for the kappa index. Full article

Interferometric synthetic aperture radar (InSAR) technology has been widely employed in the rapid monitoring of earthquakes and associated geological hazards. With the continued advancement of InSAR technology, the growing volume of satellite-acquired data has opened new avenues for applying deep learning (DL) techniques to the analysis of earthquake-induced surface deformation. Although DL holds great promise for processing InSAR data, its development progress has been significantly constrained by the absence of large-scale, accurately annotated datasets related to earthquake-induced deformation. To address this limitation, we propose an automated method for constructing deep learning training datasets by integrating the Global Centroid Moment Tensor (GCMT) earthquake catalog with Sentinel-1 InSAR observations. This approach reduces the inefficiencies and manual labor typically involved in InSAR data preparation, thereby significantly enhancing the efficiency and automation of constructing deep learning datasets for coseismic deformation. Using this method, we developed and publicly released a large-scale training dataset consisting of coseismic InSAR samples. The dataset contained 353 Sentinel-1 interferograms corresponding to 62 global earthquakes that occurred between 2015 and 2024. Following standardized preprocessing and data augmentation (DA), a large number of image samples were generated for model training. Multidimensional analyses of the dataset confirmed its high quality and strong representativeness, making it a valuable asset for deep learning research on coseismic deformation. The dataset construction process followed a standardized and reproducible workflow, ensuring objectivity and consistency throughout data generation. As additional coseismic InSAR observations become available, the dataset can be continuously expanded, evolving into a comprehensive, high-quality, and diverse training resource. It serves as a solid foundation for advancing deep learning applications in the field of InSAR-based coseismic deformation analysis. Full article