Search Results (11,333)

Populus euphratica is a critical constructive species in arid desert regions, serving as a “natural barrier” for oasis protection. The sustainable management of Populus euphratica forests is directly related to regional ecological security, and the fine identification of sparse Populus euphratica forests is essential for the conservation of natural Populus euphratica forests. Currently, most mapping studies on Populus euphratica distribution focus on the extraction of dense, contiguous Populus euphratica forests, with insufficient attention paid to the identification of sparse Populus euphratica forests. This study utilizes Gaofen-2 (GF-2) satellite imagery as the data source and takes a typical sparse Populus euphratica forests distribution area in the Tarim River Basin as the study site. It systematically evaluates the performance of nine mainstream deep learning models, including U-Net, DeepLabV3+, and SegFormer, in the task of sparse Populus euphratica forests identification. The results indicate that: (1) The false-color sample set, synthesized from near-infrared, red, and green bands, contributes to improved model accuracy. Compared to the true-color (red, green, blue bands) dataset, the average Intersection over Union (IoU) of the nine models shows a relative improvement of approximately 20%. (2) For the sparse Populus euphratica forests identification task based on the false-color dataset, four models—U-Net, U-Net++, MA-Net, and DeepLabV3+—exhibited excellent performance, with IoU exceeding 75%. (3) Using U-Net as the baseline model, this study integrated the max-pooling indices mechanism, atrous spatial pyramid pooling, and residual connection modules to construct a semantic segmentation network tailored for sparse Populus euphratica forests, named Sparse Populus euphratica Segmentation Network (SPS-Net). This model achieved an IoU of 80%, a relative improvement of approximately 6.3% over the baseline model, and demonstrated good stability in large-scale classification tests. The identification scheme for sparse Populus euphratica forests constructed using GF-2 imagery and deep learning models proposed in this study can provide effective technical support for the refined monitoring and protection of natural Populus euphratica forests. Full article

Unmanned aerial systems (UASs) and artificial intelligence (AI) allow for the effective monitoring of the plants, but it is difficult to determine the stages of cotton development in the process of irrigation gradients. In this paper, UAS images were combined with deep learning to conduct field-scale cotton phenology classification in graded drought situations. SegNet, U-Net, and DeepLabv3+ were trained on various sample sizes and tested on global accuracy (GA), mean intersection-over-union (mIoU), and mean boundary F-score (mBF). It was found that DeepLabv3+ outperformed all other methods and yielded the most uniform delineation of crop row spacing, canopy edges, and boll opening boundaries throughout the entire growing season. Under single-stage training, performance became stable at training sample sizes ≥ 960 for the seedling and squaring stages, whereas the boll and boll-opening stages required ≥ 1280; for full-season training, performance became stable when the sample size reached 4480 (GA = 0.98, mIoU = 0.95, mBF = 0.81). Cross-treatment evaluation indicated that errors were mainly concentrated between adjacent stages, with higher confusion under the 0% irrigation treatment and more stable identification results under the 90% irrigation treatment. A DAP 138 field survey (36 points) confirmed an irrigation-gradient phenological shift from boll-opening dominance at 0% irrigation to universal boll at 90% irrigation, consistent with spatial phenology maps. Overall, the proposed framework provides a cost-effective, field-scale solution to support precision irrigation management in arid cotton-growing regions. Full article

Functional near-infrared spectroscopy (fNIRS) is a valuable non-invasive modality for brain-computer interfaces (BCIs), but robust signal interpretation is challenged by the significant temporal variability of the hemodynamic response. Standard linear methods, such as Pearson correlation, often fail to capture functional connectivity when signals exhibit temporal jitter. This study validates an unsupervised Dynamic Time Warping (DTW) clustering framework to robustly identify motor networks from fNIRS data by accommodating non-linear temporal shifts. We analyzed a public fNIRS dataset (N = 30) across right-hand (RHT), left-hand (LHT), and foot tapping (FT) tasks. A robust preprocessing pipeline was implemented, including Wavelet Motion Correction and Common Average Referencing (CAR) to remove artifacts and global systemic noise. The core method involved computing Z-score normalized DTW distance matrices, followed by hierarchical clustering. To validate the framework, we benchmarked it against a standard Pearson Correlation method. Results show that the unsupervised DTW framework achieved a network identification accuracy of 53.17%, significantly outperforming the standard Pearson correlation benchmark (48.06%) with a statistically significant difference (p < 0.05). The framework successfully detected distinct, somatotopically correct modulations: superior-medial activation during foot tapping and lateralized activation during hand tapping. These findings demonstrate that unsupervised DTW clustering is a robust, data-driven approach that outperforms conventional linear methods in capturing functional networks during motor tasks, showing significant potential for next-generation asynchronous BCIs. Full article

Existing ferroresonance fault identification methods often suffer from high misclassification rates, strong threshold dependency, and insufficient noise resistance. To bridge this gap, we propose a novel ferroresonance fault recognition method based on the Markov transition field (MTF) and three-branch Gaussian clustering (TBGC). Firstly, a symplectic geometric algorithm is employed to denoise the resonance feature signal, extract effective dominant modes, and reshape the series. Secondly, the reshaped feature series is converted into a Pixel matrix image employing the MTF. Subsequently, the gray-level co-occurrence matrix (GLCM) is utilized to extract the two-dimensional texture features of MTF images corresponding to different resonance types and construct corresponding TBGC models. Finally, the overvoltage sequence to be recognized is input into the TBGC model after feature extraction, and accurate discrimination of ferroresonance types is achieved based on cosine similarity. The analysis of fault recording data indicates that this method achieves 100% discrimination accuracy in eight test cases, surpassing the comparative method (maximum accuracy of 62.5%) by 37.5%, thereby validating its effectiveness and accuracy in ferroresonance identification. Full article

Accurate and early identification of breast cancer from mammography is key to reducing breast cancer mortality, and automated analysis is challenging due to subtle lesion appearances, heterogeneous breast density, and the variance caused by modality. Standard Convolutional Neural Networks (CNNs) are excellent at capturing localized textures, whereas Vision Transformers (ViTs) capture long-range dependencies; however, both often struggle to produce a unified representation that consistently supports diagnostic decision-making. To address these limitations, this study presents a dual-stream framework integrating ConvNeXt for high-fidelity local feature extraction with Swin Transformer V2 for hierarchical global context modeling. A Bi-Directional Cross-Guidance (BDCG) mechanism is added to harmonize interactions between the two feature domains and ensure mutual information learning in the representations. Furthermore, a Prototype-Anchored Similarity Head (PASH) is used to stabilize classification using distance-based reasoning instead of using linear separation. Comprehensive experiments show the effectiveness of the proposed method using two benchmark datasets. On Dataset 1, the model achieves accuracy: 98.8%, precision: 98.7%, recall: 98.6%, and F1 score: 97.2%, outperforming existing models based on CNN, ViTs, and hybrid architectures, and provides a lower inference time (8.3 ms/image). On the more heterogeneous Dataset 2, the model maintains strong performance, with an accuracy of 97.0%, precision of 95.4%, recall of 94.8%, and F1-score of 95.1%, demonstrating its resilience to domain shift and imaging variability. These results underscore the value of structural multi-scale feature interaction and prototype-driven classification for robust mammographic analysis. The consistent performance across internal and external evaluations indicates the potential for the proposed framework to be reliably applied in computer-aided screening systems. Full article

With the widespread adoption of container technology, its shared kernel architecture has made abnormal file access behavior a key precursor to container escape and lateral attacks, necessitating precise and efficient runtime detection mechanisms. However, existing monitoring methods typically suffer from issues such as insufficient granularity in data collection, limited path semantic modeling capabilities, and low anomaly detection accuracy. To address these challenges, this paper proposes an eBPF-based method for detecting abnormal file access in containers. A lightweight kernel-level monitoring mechanism is constructed to capture access behavior in real time at the system call level, effectively enhancing both the granularity of data collection and the completeness of context. At the feature modeling layer, a multimodal path semantic representation method is designed, combining risk-layer rules and semantic vectorization strategies to enhance the hierarchical expression of path structures and improve context modeling ability. In the detection layer, an attention-enhanced autoencoder model is introduced, achieving high-precision identification of abnormal access behavior and low false-positive monitoring under unsupervised conditions through a path segment attention mechanism and weighted reconstruction loss function. Experiments in real container environments show that the proposed method achieves a recall rate of $82.0\%$, a false-positive rate of $0.79\%$, and a Matthews correlation coefficient of 0.852, significantly outperforming mainstream unsupervised detection methods such as Isolation Forest, One-Class SVM, and Local Outlier Factor. These results verify the advantages of the proposed method in terms of detection accuracy, real-time performance, and system friendliness, providing an efficient and feasible solution for enhancing the detection of unknown attacks in container runtimes. Full article

Modeling ship maneuvering dynamics presents numerous challenges, including long-term multi-step recursive error accumulation, insufficient generalization under distributed control rates, and high-frequency disturbance amplification effects. Traditional analytical models heavily rely on vessel-specific trials to characterize strongly nonlinear coupling terms and perform parameter identification, making it difficult to balance efficiency and accuracy under complex operating conditions. This paper presents a ship maneuvering-oriented integration of an invertible Koopman representation and a NeuralODE-based continuous-time predictor. The IKN reconstructs strongly coupled state spaces while enhancing representational invertibility, whereas NeuralODE directly fits the control differential equations governing ship maneuvering dynamics and supports continuous-time prediction. Experiments validate multi-rate control performance under ideal and disturbed data conditions, assessing error accumulation and extrapolation stability through long-term multi-step propagation. Evaluations utilize the KVLCC2-type L7 ship model with a 0.25 s sampling interval and a 200 s prediction horizon, validated against a multi-rate control test set. The results indicate that, compared to the baseline neural ODEs model without IKN, the normalized root mean square error (NRMSE) of state quantities decreased by 12.68% on average. In typical operational scenarios such as constant-speed emergency turns and variable-speed sine sweep maneuvers, the average state NRMSE was 7.96% lower than the LSTM model and 53.85% lower than the IKN–Koopman operator network. Noise experiments demonstrated that when introducing simulated sensor noise at 5%, 10%, and 20% into the dataset, the average state NRMSE remained at 5.98%, 8.24%, and 10.06%, respectively. This confirms the method’s stable prediction performance under varying noise intensities. Full article

Driven by the goal of sustainable energy transitions, the integration of Inverter-Interfaced Distributed Generation (IIDG) has led to a continuous decline in the accuracy of single-phase grounding fault line selection in neutral non-effectively grounded distribution networks. Protection methods based on characteristic signal injection currently struggle to balance the differentiated requirements of fault detection sensitivity and equipment safety in networks with high-penetration IIDG. To address this issue, a high-frequency equivalent circuit model of the IIDG is established. The distribution patterns of the high-frequency characteristic current (HFCC) in distribution networks under high-penetration IIDG are analyzed. Subsequently, an adaptive HFCC injection strategy is proposed, which accounts for IIDG low-voltage ride-through (LVRT) requirements, fault identification sensitivity, and equipment safety constraints. Based on the amplitude and phase differences in the HFCC between faulty and healthy feeders, a fault line selection criterion is established. Consequently, an adaptive injection-based protection method for single-phase grounding fault is developed, considering the impact of high-penetration IIDG. Simulation results demonstrate that the proposed method accurately identifies the faulty feeder under various fault locations, transition resistances, and quantities of integrated IIDG units. The results further confirm the high adaptability and reliability of the method, thereby providing a robust technical foundation for the safe, reliable, and sustainable operation of modern power grids. Full article

Ants are vital bioindicators that contribute to soil health and food webs, making accurate identification essential for biodiversity monitoring and conservation. However, traditional taxonomic methods are time-consuming and require specialized expertise, limiting large-scale data collection and public participation. This paper presents AntID_APP, a web-based application designed to support citizen scientists in Taiwan by enabling real-time, image-based detection and the identification of native ant genera. Fine-tuned YOLO models first detect ants in user-uploaded images and then classify them at the genus level. The models were trained on a curated dataset of 60,429 open-access images from iNaturalist, covering 54 native ant species. To ensure robustness in real-world conditions, we applied targeted data augmentation and evaluated multiple YOLO versions (v9–v12). The best-performing model achieved a mean Average Precision (mAP50: 0.935–0.948, mAP50-95: 0.777–0.807) for the detection task, followed by accurate genus-level identification. The application features an intuitive interface and a lightweight asynchronous server architecture, allowing users to upload images and receive both visual detection results (bounding boxes) and genus predictions efficiently. By combining high accuracy with accessibility, AntID_APP offers a scalable solution for biodiversity monitoring and public engagement in ecological research. Full article

Accurate cotton boll identification and boll number estimation from UAV imagery are essential for large-scale yield prediction and precision management, yet severe leaf occlusion and complex canopy backgrounds often hinder robust performance. Here, UAV RGB images were acquired 3 days before defoliant application and at 3, 6, 9, 12, 15, and 18 days after defoliation. Cotton bolls were extracted using Mahalanobis distance, a support vector machine, and a neural network. Boll number was then estimated using an improved random forest model with multi-feature fusion. Across all defoliation stages, the NN produced the most accurate and stable boll extraction, achieving a maximum Kappa of 0.914, an overall accuracy of 95.77%, and an F1 score of 0.96. Extraction accuracy increased rapidly from 3 to 9 days after application and stabilized from 12 to 18 days. For boll number estimation, fusing the boll pixel ratio with color indices and texture features improved accuracy and consistency over time; the best performance was obtained at 18 days after application (R² = 0.7264; rRMSE = 4.9%). Overall, imagery acquired 15–18 days after defoliation provided the most reliable estimation window, supporting operational pre-harvest assessment and harvest-timing decisions. Full article

Early-stage fungal contamination in maize kernels is difficult to identify visually and it can cause severe quality and safety risks during storage and transportation. Short-wave infrared (SWIR) hyperspectral imaging offers a rapid, non-destructive approach by capturing chemical information related to water, proteins, and lipids. This study investigates the early detection and classification of Gibberella zeae contamination in maize kernels using SWIR hyperspectral imaging combined with machine learning. Two maize varieties were artificially inoculated and cultured under controlled conditions, followed by hyperspectral data collection over six contamination stages. Various preprocessing techniques including standard normal variate (SNV), second derivative (SD), multiplicative scatter correction (MSC), and derivatives were evaluated to enhance data quality. Feature wavelength selection was performed using successive projections algorithm (SPA), competitive adaptive reweighted sampling (CARS), and uninformative variable elimination (UVE), significantly reducing redundancy and improving classification performance. Multiple models, including linear discriminant analysis (LDA), multilayer perceptron (MLP), support vector machine (SVM), a convolutional neural network (CNN), long short-term memory (LSTM) network, and a hybrid architecture Transformer that integrated a CNN, a LSTM network, and a Transformer (abbreviated as CLT), were constructed for both binary (healthy vs. contaminated) and multiclass classification tasks. Specifically, the multiclass task consisted of six contamination stages corresponding to contamination time from Day 0 to Day 5. The best binary classification task accuracy of 100% was achieved using SNV-preprocessed data with the MLP model. For multiclass classification task, the SD-preprocessed LDA model reached a test accuracy of 92.56%. Combined with appropriate preprocessing, feature selection and modeling, these results demonstrate that hyperspectral imaging is a powerful tool for the non-destructive, early-stage identification of fungal contamination in maize kernels, offering strong support for food safety and quality monitoring. Full article

Accurate and rapid identification of bacterial species is essential for public health, clinical diagnostics, and environmental monitoring. Although Raman spectroscopy offers a powerful, non-invasive alternative, reliance solely on spectral data often fails to distinguish species with highly similar signatures, particularly when the discriminating features are subtle. This difficulty is frequently compounded by a lack of integrated biological prior knowledge, which can hinder model performance. To address these challenges, we introduce BactoRamanBioNet, a novel multimodal neural network architecture. Our model employs a synergistic approach that utilizes a ResNet-Transformer architecture to capture complex spectral patterns and a CLIP text encoder to incorporate descriptive biological information, thereby enabling highly accurate multimodal classification of bacterial species. Empirical results demonstrate that BactoRamanBioNet achieves a classification accuracy of 98.2% and an F1-score of 98.0%. This performance surpasses the current state-of-the-art deep learning model, ResNet-1D, by 2.4% in accuracy and 2.0% in F1-score. Moreover, our model outperforms traditional classifiers, such as Support Vector Machine (SVM) and Random Forest (RF), by 9.8% and 7.9% in accuracy, respectively, while also exhibiting significant improvements in precision and recall. By establishing a new benchmark in performance and robustness, BactoRamanBioNet offers a powerful and reliable framework for automated microbiological analysis, paving the way for next-generation diagnostic systems. Full article

Loss of skeletal muscle mass in cancer cachexia is associated with poorer survival, reduced treatment tolerance, and diminished quality of life. Routine oncology computed tomography (CT) can yield skeletal muscle area (SMA) and skeletal muscle index (SMI) for early cachexia assessment and prognostication, but manual annotation is labor intensive and existing automated tools often show inconsistent reliability. We developed SMAART-AI (Skeletal Muscle Assessment—Automated and Reliable Tool based on AI), a fully automated pipeline that localizes the third lumbar (L3) vertebral level, segments skeletal muscle, and quantifies prediction uncertainty to flag potentially unreliable outputs. Performance and reliability were evaluated across gastroesophageal, pancreatic, colorectal, and ovarian cancer cohorts, benchmarking against expert annotations and existing tools. SMAART-AI achieved a Dice score of 97.80% ± 0.93% in gastroesophageal cancer and a median SMA deviation of 2.48% from expert annotations across pancreatic, colorectal, and ovarian cohorts. Uncertainty scores correlated strongly with prediction error, enabling identification of high-error cases to support trustworthy deployment. Integrating the SMA/SMI with clinical features and body mass index (BMI) improved survival prediction (concordance index was +2.19% for colorectal, +9.82% for pancreatic, and +2.58% for ovarian cancer) and supported cachexia detection (70.00% accuracy; F1 80.00%). Overall, SMAART-AI provides an uncertainty-aware, clinically translatable framework for scalable CT-based muscle assessment and improved oncologic prognostication. Full article

Land surface bidirectional reflectance distribution functions (BRDF) are critical for quantitative remote sensing but are significantly constrained by scale effects, limiting the interoperability of multi-resolution data and the accuracy of quantitative inversion, thereby rendering the investigation of BRDF multi-scale effects increasingly urgent. This study utilized UAV (Unmanned Aerial Vehicle)-based multi-angular observations and the RPV model to retrieve the BRDF of typical land covers, employing the Window Averaging Method to simulate multi-scale responses and systematically investigate the relationship between BRDF characteristics and spatial scale. The results indicate the following key findings: (1) The RPV (Rahman–Pinty–Verstraete) model demonstrated high robustness and inversion accuracy, yielding RMSE (Root Mean Square Error) below 0.06 and RRMSE (Relative RMSE) below 25% across all land covers, with the 840 nm band exhibiting superior performance. (2) Significant spatial scale effects were observed, where BRDF characteristics varied distinctively with scale but eventually stabilized at specific thresholds; specifically, the stabilization scales were identified as 1.3 m for bare soil, 1.5 m for tea plantations, 1 m for rice, and 2 m for forests. (3) The scale evolution of BRDF features exhibited a parallel trend with spatial heterogeneity, a correlation that enables the quantitative identification of optimal observation scales for different land cover types. Full article

In the practice of enterprise data security governance, document AI has emerged as a mission-critical component that seeks to underpin the prevention of document leakage via automatic accurate classification and identification of sensitive content. Arising from this, a need to bring document classification benchmark closer to real-world engineering applications is highlighted. This paper identifies the lack of public datasets for native multi-modal hybrid document classification and, accordingly, proposes the dataset DocCLS_NMMH (Native Multi-Modal Hybrid Document Classification) along with its out-of-distribution (OOD) test subset. An experimental study on the proposed dataset demonstrates that current benchmarks have become irrelevant and need to be updated to evaluate native multi-modal hybrid documents. Meanwhile, accuracy degradation in heterogeneous documents and few-shot scenarios is assessed, as all of these are prevalent in the practice. The experimental results demonstrate that LayoutLM achieves a state-of-the-art (SOTA) performance with 98.66% accuracy on DocCLS_NMMH, with only approximately 7% accuracy degradation on its OOD test subset, while training-free models (Qwen2.5-VL-32B and Gemma3-27B) consistently achieve over 95% accuracy across the full dataset. The SOTA performance of these models on our benchmark provides an effective guidance for model selection in real engineering applications. Full article