Search Results (7,085)

When faced with diverse types of image tampering and image quality degradation in real-world scenarios, traditional image tampering localization methods often struggle to balance boundary accuracy and robustness. To address these issues, this paper proposes a symmetric guided dual-branch image tampering localization network—FENet (Fusion-Enhanced Network)—that integrates adaptive feature fusion and edge attention mechanisms. This method is based on a structurally symmetric dual-branch architecture, which extracts RGB semantic features and SRM noise residual information to comprehensively capture the fine-grained differences in tampered regions at the visual and statistical levels. To effectively fuse different features, this paper designs a self-calibrating fusion module (SCF), which introduces a content-aware dynamic weighting mechanism to adaptively adjust the importance of different feature branches, thereby enhancing the discriminative power and expressiveness of the fused features. Furthermore, considering that image tampering often involves abnormal changes in edge structures, we further propose an edge-aware coordinate attention mechanism (ECAM). By jointly modeling spatial position information and edge-guided information, the model is guided to focus more precisely on potential tampering boundaries, thereby enhancing its boundary detection and localization capabilities. Experiments on public datasets such as Columbia, CASIA, and NIST16 demonstrate that FENet achieves significantly better results than existing methods. We also analyze the model’s performance under various image quality conditions, such as JPEG compression and Gaussian blur, demonstrating its robustness in real-world scenarios. Experiments in Facebook, Weibo, and WeChat scenarios show that our method achieves average F1 scores that are 2.8%, 3%, and 5.6% higher than those of existing state-of-the-art methods, respectively. Full article

This study analyses algorithmic biases in AI-generated images, focusing on aesthetic violence, gender stereotypes, and weight discrimination. By examining images produced by the DALL-E Nature and Flux 1 systems, it becomes evident how these tools reproduce and amplify hegemonic beauty standards, excluding bodily diversity. Likewise, gender representations reinforce traditional roles, sexualising women and limiting the presence of non-normative bodies in positive contexts. The results show that training data and the algorithms used significantly influence these trends, perpetuating exclusionary visual narratives. The research highlights the need to develop more inclusive and ethical AI models, with diverse data that reflect the plurality of bodies and social realities. The study concludes that artificial intelligence (AI), far from being neutral, actively contributes to the reproduction of power structures and inequality, posing an urgent challenge for the development and regulation of these technologies. Full article

The rapid development of deepfake technologies has led to the widespread proliferation of facial image forgeries, raising significant concerns over identity theft and the spread of misinformation. Although recent dual-domain detection approaches that integrate spatial and frequency features have achieved noticeable progress, they still suffer from limited sensitivity to local forgery regions and inadequate interaction between spatial and frequency information in practical applications. To address these challenges, we propose a novel forgery-aware guided spatial–frequency feature fusion network. A lightweight U-Net is employed to generate pixel-level saliency maps by leveraging structural symmetry and semantic consistency, without relying on ground-truth masks. These maps dynamically guide the fusion of spatial features (from an improved Swin Transformer) and frequency features (via Haar wavelet transforms). Cross-domain attention, channel recalibration, and spatial gating are introduced to enhance feature complementarity and regional discrimination. Extensive experiments conducted on two benchmark face forgery datasets, FaceForensics++ and Celeb-DFv2, show that the proposed method consistently outperforms existing state-of-the-art techniques in terms of detection accuracy and generalization capability. The future work includes improving robustness under compression, incorporating temporal cues, extending to multimodal scenarios, and evaluating model efficiency for real-world deployment. Full article

Accurate diagnosis of knee joint injuries from magnetic resonance (MR) images is critical for patient care. Background/Objectives: While deep learning has advanced 3D MR image analysis, its reliance on extensive labeled datasets is a major hurdle for diverse knee pathologies. Few-shot learning (FSL) addresses this by enabling models to classify new conditions from minimal annotated examples, often leveraging knowledge from related tasks. However, creating robust 3D FSL frameworks for varied knee injuries remains challenging. Methods: We introduce MedNet-FS, a 3D FSL framework that effectively classifies knee injuries by utilizing domain-specific pre-trained weights and generalized end-to-end (GE2E) loss for discriminative embeddings. Results: MedNet-FS, with knee-MRI-specific pre-training, significantly outperformed models using generic or other medical pre-trained weights and approached supervised learning performance on internal datasets with limited samples (e.g., achieving an area under the curve (AUC) of 0.76 for ACL tear classification with k = 40 support samples on the MRNet dataset). External validation on the KneeMRI dataset revealed challenges in classifying partially torn ACL (AUC up to 0.58) but demonstrated promising performance for distinguishing intact versus fully ruptured ACLs (AUC 0.62 with k = 40). Conclusions: These findings demonstrate that tailored FSL strategies can substantially reduce data dependency in developing specialized medical imaging tools. This approach fosters rapid AI tool development for knee injuries and offers a scalable solution for data scarcity in other medical imaging domains, potentially democratizing AI-assisted diagnostics, particularly for rare conditions or in resource-limited settings. Full article

Formation MicroScanner Image (FMI) technology is a key method for identifying fractured reservoirs and optimizing oil and gas exploration, but its inherent insufficient resolution severely constrains the fine characterization of geological features. This study innovatively applies a Super-Resolution Generative Adversarial Network (SRGAN) to the super-resolution reconstruction of FMI logging image to address this bottleneck problem. By collecting FMI logging image of glutenite from a well in Xinjiang, a training set containing 24,275 images was constructed, and preprocessing strategies such as grayscale conversion and binarization were employed to optimize input features. Leveraging SRGAN’s generator-discriminator adversarial mechanism and perceptual loss function, high-quality mapping from low-resolution FMI logging image to high-resolution images was achieved. This study yields significant results: in RGB image reconstruction, SRGAN achieved a Peak Signal-to-Noise Ratio (PSNR) of 41.39 dB, surpassing the optimal traditional method (bicubic interpolation) by 61.6%; its Structural Similarity Index (SSIM) reached 0.992, representing a 34.1% improvement; in grayscale image processing, SRGAN effectively eliminated edge blurring, with the PSNR (40.15 dB) and SSIM (0.990) exceeding the suboptimal method (bilinear interpolation) by 36.6% and 9.9%, respectively. These results fully confirm that SRGAN can significantly restore edge contours and structural details in FMI logging image, with performance far exceeding traditional interpolation methods. This study not only systematically verifies, for the first time, SRGAN’s exceptional capability in enhancing FMI resolution, but also provides a high-precision data foundation for reservoir parameter inversion and geological modeling, holding significant application value for advancing the intelligent exploration of complex hydrocarbon reservoirs. Full article

Recently, a network based on selective state space models (SSMs), Mamba, has emerged as a research focus in hyperspectral image (HSI) classification due to its linear computational complexity and strong long-range dependency modeling capability. Originally designed for 1D causal sequence modeling, Mamba is challenging for HSI tasks that require simultaneous awareness of spatial and spectral structures. Current Mamba-based HSI classification methods typically convert spatial structures into 1D sequences and employ various scanning patterns to capture spatial dependencies. However, these approaches inevitably disrupt spatial structures, leading to ineffective modeling of complex spatial relationships and increased computational costs due to elongated scanning paths. Moreover, the lack of neighborhood spectral information utilization fails to mitigate the impact of spatial variability on classification performance. To address these limitations, we propose a novel model, Dual-Aware Discriminative Fusion Mamba (DADFMamba), which is simultaneously aware of spatial-spectral structures and adaptively integrates discriminative features. Specifically, we design a Spatial-Structure-Aware Fusion Module (SSAFM) to directly establish spatial neighborhood connectivity in the state space, preserving structural integrity. Then, we introduce a Spectral-Neighbor-Group Fusion Module (SNGFM). It enhances target spectral features by leveraging neighborhood spectral information before partitioning them into multiple spectral groups to explore relations across these groups. Finally, we introduce a Feature Fusion Discriminator (FFD) to discriminate the importance of spatial and spectral features, enabling adaptive feature fusion. Extensive experiments on four benchmark HSI datasets demonstrate that DADFMamba outperforms state-of-the-art deep learning models in classification accuracy while maintaining low computational costs and parameter efficiency. Notably, it achieves superior performance with only 30 training samples per class, highlighting its data efficiency. Our study reveals the great potential of Mamba in HSI classification and provides valuable insights for future research. Full article

The integration of HS-SPME-GC/MS and UPLC-MS/MS techniques enabled the profiling of volatile organic compounds (VOCs) and amino acids (AAs) in 18 crabapple flower cultivars, facilitating the development of a novel VOC–AA model. Among the 51 identified VOCs, benzyl alcohol, benzaldehyde, and ethyl benzoate were predominant, categorizing cultivars into fruit-almond, fruit-sweet, and mixed types. The amino acids, namely glutamic acid (Glu), asparagine (Asn), aspartic acid (Asp), serine (Ser), and alanine (Ala) constituted 83.6% of the total AAs identified. Notably, specific amino acids showed positive correlations with key VOCs, suggesting a metabolic regulatory mechanism. The Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) model, when combined with volatile organic compounds (VOCs) and amino acid profiles, enabled more effective aroma type classification, providing a robust foundation for further studies on aroma mechanisms and targeted breeding. Full article

Background/Objectives: The oral cavity represents a convenient route of administration for drugs that exhibit significant hepatic first-pass extraction. In this study, the mucosal permeation properties of selected active pharmaceutical ingredients (APIs) incorporated into oral cavity drug products that are approved by the U.S. Food and Drug Administration were quantified using the human-derived sublingual HO-1-u-1 and buccal EpiOral™ in vitro tissue models. Methods: Epithelial barrier properties were monitored using propranolol and Lucifer Yellow as prototypic transcellular and paracellular markers. APIs were dissolved in artificial saliva, pH 6.7, and transepithelial flux from the apical to the basolateral compartment was quantified using HPLC. Results: Apparent permeability coefficients (Papp) calculated for these APIs in the sublingual HO-1-u-1 tissue model varied from Papp = 2.72 ± 0.06 × 10⁻⁵ cm/s for asenapine to Papp = 6.21 ± 2.60 × 10⁻⁵ cm/s for naloxone. In contrast, the buccal EpiOral™ tissue model demonstrated greater discrimination power in terms of permeation properties for the same APIs, with values ranging from Papp = 3.31 ± 0.83 × 10⁻⁷ cm/s for acyclovir to Papp = 2.56 ± 0.68 × 10⁻⁵ cm/s for sufentanil. The tissue-associated dose fraction recovered at the end of the transport experiment was significantly increased in the buccal EpiOral™ tissue model, reaching up to 8.5% for sufentanil. Conclusions: Experimental permeation data collected for selected APIs in FDA-approved oral cavity products will serve as a training set to aid the development of predictive computational models for improving algorithms that describe drug absorption from the oral cavity. Following a robust in vitro–in vivo correlation analysis, it is expected that such innovative in silico modeling strategies will the accelerate development of generic oral cavity products by facilitating the utility of model-integrated evidence to support decision making in generic drug development and regulatory approval. Full article

Mass spectrometric (MS) fingerprinting coupled with chemometrics for the detection of plants in plant mixtures is sparsely researched. This paper aims to check its value for herbal adulteration concerning plants with slimming as an indication. Moreover, it is among the first to exploit the full three-dimensional dataset (i.e., time × intensity × mass) obtained with liquid chromatography hyphenated with MS for herbal fingerprinting purposes. The MS parameters were optimized to achieve highly specific fingerprints. Trituration’s (total 55), blanks (total 11) and reference plants were injected in the MS system to generate the dataset. The dataset was complex and humongous, necessitating the application of compression techniques. After compression, Partial Least Squares-Discriminant Analysis (PLS-DA) was performed to generate models validated for accuracy using cross-validation and an external test set. Confusion matrices were constructed to provide insight into the modeling predictions. A complimentary evaluation between data obtained using a previously developed Diode Array Detection (DAD) method and the MS data was performed by data fusion techniques and newly generated models. The fused dataset models were comparable to MS models. For ease of application, MS modeling was deemed to be superior. The future market studies would adopt MS modeling as the preferred choice. A proof of concept was carried out on 10 real-life samples obtained from illegal sources. The results indicated the need for stronger monitoring of (illegal) plant food supplements entering the market, especially via the internet. Full article

The agriculture sector is increasingly challenged to maintain productivity and sustainability amidst environmental, marketplace, and geopolitical pressures. While precision agriculture enhances physical production, the financial resilience of agricultural firms has been understudied. In this study, machine learning (ML) methods, including logistic regression (LR), decision trees (DTs), and artificial neural networks (ANNs), are employed to predict the bankruptcy risk for Central and Eastern European (CEE) farming firms. All models consistently showed high performance, with AUC values exceeding 0.95. DTs had the highest overall accuracy (95.72%) and F1 score (0.9768), LR had the highest recall (0.9923), and ANNs had the highest discrimination power (AUC = 0.960). Visegrad, Balkan, Baltic, and Eastern Europe subregional models featured economic and structural heterogeneity, reflecting the need for local financial risk surveillance. The results support the development of AI-based early warning systems for agricultural finance, enabling smarter decision-making, regional adaptation, and enhanced sustainability in the sector. Full article

Infrared small-target detection remains a critical challenge in military reconnaissance, environmental monitoring, forest-fire prevention, and search-and-rescue operations, owing to the targets’ extremely small size, sparse texture, low signal-to-noise ratio, and complex background interference. Traditional convolutional neural networks (CNNs) struggle to detect such weak, low-contrast objects due to their limited receptive fields and insufficient feature extraction capabilities. To overcome these limitations, we propose a Multi-Scale Edge-Aware Convolution (MEAC) module that enhances feature representation for small infrared targets without increasing parameter count or computational cost. Specifically, MEAC fuses (1) original local features, (2) multi-scale context captured via dilated convolutions, and (3) high-contrast edge cues derived from differential Gaussian filters. After fusing these branches, channel and spatial attention mechanisms are applied to adaptively emphasize critical regions, further improving feature discrimination. The MEAC module is fully compatible with standard convolutional layers and can be seamlessly embedded into various network architectures. Extensive experiments on three public infrared small-target datasets (SIRSTD-UAVB, IRSTDv1, and IRSTD-1K) demonstrate that networks augmented with MEAC significantly outperform baseline models using standard convolutions. When compared to eleven mainstream convolution modules (ACmix, AKConv, DRConv, DSConv, LSKConv, MixConv, PConv, ODConv, GConv, and Involution), our method consistently achieves the highest detection accuracy and robustness. Experiments conducted across multiple versions, including YOLOv10, YOLOv11, and YOLOv12, as well as various network levels, demonstrate that the MEAC module achieves stable improvements in performance metrics while slightly increasing computational and parameter complexity. These results validate the MEAC module’s significant advantages in enhancing the detection of small and weak objects and suppressing interference from complex backgrounds. These results validate MEAC’s effectiveness in enhancing weak small-target detection and suppressing complex background noise, highlighting its strong generalization ability and practical application potential. Full article

The quality and freshness of fruits and vegetables are critical factors in consumer acceptance and are significantly affected during transport and storage. This study aimed to evaluate the quality of greenhouse-grown tomatoes stored for 24 days by combining non-destructive image analysis, spectrophotometric assays (including total phenolic content and antioxidant and antiradical activity assessments), and attenuated total reflectance–Fourier transform infrared (ATR-FTIR) spectroscopy. Additionally, water activity, moisture content, total soluble solids, texture, and color were evaluated. Most physicochemical changes occurred between days 14 and 17, without major impact on overall fruit quality. A progressive transition in peel hue from orange to dark orange, and increased surface irregularity of their textural image were noted. Moreover, the combined use of instrumental and image analyses results via multivariate analysis allowed the clear discrimination of tomatoes according to storage days. In this sense, tomato samples were effectively classified by ATR-FTIR spectral bands, linked to carotenoids, phenolics, and polysaccharides. Machine learning (ML) models, including Random Forest and Gradient Boosting, were trained on image-derived features and accurately predicted shelf life and quality traits, achieving R² values exceeding 0.9. The findings demonstrate the effectiveness of combining imaging, spectroscopy, and ML for non-invasive tomato quality monitoring and support the development of predictive tools to improve postharvest handling and reduce food waste. Full article

Background/Objectives: Non-alcoholic fatty liver disease (NAFLD) represents significant health challenges, especially among patients with chronic hepatitis B (CHB). This study uses machine learning models to predict NAFLD in patients with inactive CHB. It builds on previous research by employing classification algorithms to analyze demographic, clinical, and laboratory data to identify NAFLD predictors. Methods: A single-center cross-sectional study was conducted, including 450 inactive CHB patients from Sultan Qaboos University Hospital. Five ML models were developed: Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM), and Multi-Layer Perceptron (MLP). Results: The prevalence of NAFLD was 50.22%. Among the machine learning models, Random Forest achieved the highest performance with an ROC AUC of 0.983 (95% CI: 0.952–0.999), followed by XGBoost at 0.977 (95% CI: 0.938–0.999) and MLP at 0.963 (95% CI: 0.915–0.995). SVM also showed strong performance with an AUC of 0.949 (95% CI: 0.897–0.985), while Logistic Regression demonstrated comparatively lower discrimination with an AUC of 0.886 (95% CI: 0.799–0.952). Key predictive features identified included platelet count, low-density lipoprotein (LDL), hemoglobin, and alanine aminotransferase (ALT). Logistic Regression highlighted platelet count as the most significant negative predictor, while LDL and ALT were positive contributors. Conclusions: This study shows the utility of ML in improving the identification and management of NAFLD in CHB patients, enabling targeted interventions. Future research should expand on these findings, integrating genetic and lifestyle factors to enhance predictive accuracy across diverse populations. Full article

Growing concern over anthropogenic underwater noise, highlighted by initiatives like the Marine Strategy Framework Directive (MSFD) and its Technical Group on Underwater Noise (TG Noise), emphasizes regions like the Western Black Sea, where increasing activities threaten marine habitats. This region is experiencing rapid growth in maritime traffic and resource exploitation, which is intensifying concerns over the noise impacts on its unique marine habitats. While machine learning offers promising solutions, a research gap persists in comprehensively evaluating diverse ML models within an integrated framework for complex underwater acoustic data, particularly concerning real-world data limitations like class imbalance. This paper addresses this by presenting a multi-faceted framework using passive acoustic monitoring (PAM) data from fixed locations (50–100 m depth). Acoustic data are processed using advanced signal processing (broadband Sound Pressure Level (SPL), Power Spectral Density (PSD)) for feature extraction (Mel-spectrograms for deep learning; PSD statistical moments for classical/unsupervised ML). The framework evaluates Convolutional Neural Networks (CNNs), Random Forest, and Support Vector Machines (SVMs) for noise event classification, alongside Gaussian Mixture Models (GMMs) for anomaly detection. Our results demonstrate that the CNN achieved the highest classification accuracy of 0.9359, significantly outperforming Random Forest (0.8494) and SVM (0.8397) on the test dataset. These findings emphasize the capability of deep learning in automatically extracting discriminative features, highlighting its potential for enhanced automated underwater acoustic monitoring. Full article

In recent years, heavy metal contamination in agricultural products has become a growing concern in the field of food safety. Copper (Cu) stress in crops not only leads to significant reductions in both yield and quality but also poses potential health risks to humans. This study proposes an efficient and precise non-destructive detection method for Cu stress in oilseed rape, which is based on hyperspectral false-color image construction using principal component analysis (PCA). By comprehensively capturing the spectral representation of oilseed rape plants, both the one-dimensional (1D) spectral sequence and spatial image data were utilized for multi-class classification. The classification performance of models based on 1D spectral sequences was compared from two perspectives: first, between machine learning and deep learning methods (best accuracy: 93.49% vs. 96.69%); and second, between shallow and deep convolutional neural networks (CNNs) (best accuracy: 95.15% vs. 96.69%). For spatial image data, deep residual networks were employed to evaluate the effectiveness of visible-light and false-color images. The RegNet architecture was chosen for its flexible parameterization and proven effectiveness in extracting multi-scale features from hyperspectral false-color images. This flexibility enabled RegNetX-6.4GF to achieve optimal performance on the dataset constructed from three types of false-color images, with the model reaching a Macro-Precision, Macro-Recall, Macro-F₁, and Accuracy of 98.17%, 98.15%, 98.15%, and 98.15%, respectively. Furthermore, Grad-CAM visualizations revealed that latent physiological changes in plants under heavy metal stress guided feature learning within CNNs, and demonstrated the effectiveness of false-color image construction in extracting discriminative features. Overall, the proposed technique can be integrated into portable hyperspectral imaging devices, enabling real-time and non-destructive detection of heavy metal stress in modern agricultural practices. Full article