Search Results (3,379)

Cataract diagnosis requires a comprehensive dilated examination by an ophthalmologist using a slit lamp; there is currently no effective means to objectively screen for cataracts in the community using portable devices without dilation. We hypothesized that it would be possible to predict cataract severity using deep learning on images taken using a portable smartphone-based slit lamp prototype, with and without dilation. In this prospective cross-sectional pilot study, slit lamp images were captured from eligible patients with cataracts in a tertiary clinic using a portable slit lamp prototype attached to a smartphone. The Pentacam nuclear staging score (PNS, Pentacam^®, Oculus, Inc., Arlington, WA, USA) was taken from the dilated pupils and served as ground truth. A transformer prototypical network with the Swin transformer on the images was trained to assign the class label corresponding to the highest predicted probability. Heat maps were generated based on attribution masks to identify the anatomical areas of concern. A total of 1900 images from 198 eyes of 99 patients were captured. The average age was 65.3 ± 10.4 years (range, 41.0 to 88.0 years) and the average PNS score was 1.57 ± 0.81 (range, 0 to 4). The model achieved an average accuracy of 81.25% and 74.38% for undilated and dilated eyes, respectively. Heat map visualization using the integrated gradient method successfully identified the anatomical area of interest in certain images. This study suggests the possibility of estimating cataract density using a portable smartphone slit lamp device without dilation. Further work is under way to validate this technique in a larger and more diverse group of eyes with cataracts. Full article

Accurate flood detection is critical for disaster response, yet the scarcity of diverse annotated datasets hinders robust model development. Existing resources typically suffer from limited geographic scope and insufficient annotation granularity, restricting the generalization capabilities of computer vision methods. To bridge this gap, we introduce AIFloodSense, a comprehensive evaluation benchmark designed to advance domain-generalized Artificial Intelligence for climate resilience. The dataset comprises 470 high-resolution aerial images capturing 230 distinct flood events across 64 countries and six continents. Unlike prior benchmarks, AIFloodSense ensures exceptional global diversity and temporal relevance (2022–2024), supporting three complementary tasks: (i) Image Classification, featuring novel sub-tasks for environment type, camera angle, and continent recognition; (ii) Semantic Segmentation, providing precise pixel-level masks for flood, sky, buildings, and background; and (iii) Visual Question Answering (VQA), enabling natural language reasoning for disaster assessment. We provide baseline benchmarks for all tasks using state-of-the-art architectures, demonstrating the dataset’s complexity and its utility in fostering robust AI tools for environmental monitoring. Crucially, we show that despite its compact size, AIFloodSense enables better generalization on external test sets than much larger alternatives, validating the premise that rigorous diversity is more effective than scale for training robust flood detection models, and is made publicly available to accelerate further research in the field. Full article

Emotion classification is a fundamental task in affective computing, with applications in human–computer interaction, mental health monitoring, and social media analysis. Although most existing methods formulate it as a flat classification problem, emotional expressions are inherently structured and grounded in semantic roles such as the emotion cue, stimulus, experiencer, and target. However, the relative contribution of these roles to emotion inference has not been systematically examined. Unlike prior models, we propose RolEmo, a role-aware framework for emotion classification that explicitly incorporates semantic role information. The framework employs a controlled role-masking strategy to analyze the contribution of individual roles, augments textual representations with external commonsense knowledge to capture implicit affective context, and applies supervised contrastive learning to structure the embedding space by bringing emotionally similar instances closer while separating opposing ones. We evaluate RolEmo on three benchmark datasets annotated with semantic roles. Experimental results demonstrate that RolEmo outperforms the strongest baseline across three datasets by up to $16.4\%$, $25.8\%$, and $23.2\%$ in the Full Text, Only Role, and Without Role settings, respectively. The analysis further indicates that the cue and stimulus roles provide the most reliable signals for emotion classification, with their removal causing performance drops of up to $6.2\%$ in macro f1-score, while experiencer and target roles exhibit more variable effects. These findings highlight the importance of structured semantic modeling and commonsense reasoning for robust and interpretable emotion understanding. Full article

Accurate extraction of crop planting structures is important for crop area and yield estimation, but complex and fragmented cropping patterns with overlapping phenology in the Hetao Irrigation District hinder reliable crop discrimination. This study proposes a hierarchical workflow that integrates vegetation masking with multi-source feature optimization for crop mapping. First, dual-temporal Sentinel-2 imagery (May and August) is used to generate a vegetation region-of-interest(ROI) mask via Otsu thresholding applied to the Normalized Difference Vegetation Index (NDVI), combined with pixel-wise maximum-value fusion to reduce phenology-driven omissions and background interference. Second, within the vegetation mask, Sentinel-2 spectral, vegetation-index, and texture features are combined with Sentinel-1 synthetic aperture radar (SAR) backscatter and SAR texture features to construct a multi-source feature set. Random Forest(RF) feature-importance ranking is used to select an effective feature subset, and four classifiers (RF, support vector machine (SVM), eXtreme Gradient Boosting (XGBoost), and convolutional neural network (CNN)) are compared under the same training/validation setting. The vegetation extraction achieves an overall accuracy of 91% (Kappa = 0.80). Using Sentinel-2 features only, the optimized subset with CNN attains the best performance (overall accuracy = 95%, Kappa = 0.93). Adding Sentinel-1 SAR texture features provides an additional improvement (overall accuracy = 96%, Kappa = 0.94), particularly for classes prone to confusion in fragmented plots. Area proportions derived from the final map are consistent with statistical yearbook data (percentage errors: maize 3.45%, sunflower 2.66%, wheat 0.11%, tomato 0.92%) under the study conditions. This workflow supports practical crop-structure monitoring in complex irrigation districts. Full article

We present frequency- and magnetic field-dependent measurements of the complex dielectric permittivity ε*(f, H) of a kerosene-based ferrofluid, containing Mn₀.₆Fe₀.₄Fe₂O₄ nanoparticles, over 0.8–5 GHz and static fields up to ~91 kA/m. The imaginary part, ${\epsilon }_F^{″}$, shows a peak at a characteristic frequency that shifts towards higher frequencies with increasing H, revealing a magnetic field-dependent relaxation process, interpreted using the Maxwell–Wagner–Sillars model. The dielectrophoretic extraction of nanoparticles was evaluated via the squared electric field gradient, and a threshold, ${\left(\nabla E^2\right)}_{\min }$, dependent on particle size was determined. Below that threshold, Brownian forces dominate, so the ferrofluid acts as a homogeneous dielectric. For this case, the Clausius–Mossotti factor (CM) was calculated for ferrofluid droplets in air and in water as a function of frequency and magnetic field. In air, CM exhibits modest but systematic magnetic field dependence, indicating a magnetically modulated dielectric response at GHz frequencies. In contrast, when water is used as the reference medium, CM remains negative and essentially independent of H across the entire frequency range, suggesting that the high permittivity of water masks the magneto-dielectric effects in the ferrofluid. These findings provide insight into the interplay between the magnetic field and the permittivity of ferrofluids, with implications for high-frequency applications. Moreover, using a λ/4 antenna connected to a network analyzer, the existence of the dielectrophoretic force acting on a ferrofluid-impregnated textile thread at microwave frequencies was experimentally demonstrated. Full article

Machining process route planning (MPRP) is vital for autonomous manufacturing yet remains challenging under complex, multi-dimensional engineering constraints. This paper proposes an attention-augmented deep reinforcement learning (DRL) framework to achieve intelligent process orchestration. First, an Optional Process Attribute Adjacency Graph (OPAAG) is established to formally model the “feature–process–resource–constraint” coupling, enhancing the agent’s perception of manufacturing semantics. The architecture synergistically integrates Graph Attention Networks (GAT) to perceive spatial benchmark dependencies and a Transformer-based encoder to capture sequential resource correlations within variable-length machining chains. Furthermore, a dynamic action masking mechanism is integrated to guarantee a 100% constraint satisfaction rate during both training and inference stages. Experimental evaluations across diverse part geometries demonstrate that the proposed method offers significant advantages in cost optimization, inference efficiency, and topological stability compared to traditional heuristic algorithms and standard DRL models. By effectively distilling the search space and maintaining action feasibility, the framework provides an efficient and robust solution for autonomous process planning in complex industrial scenarios. Full article

Control charts are widely used in the industrial world to monitor the average and variability of production processes. Max-Half-Mchart is a multivariate control chart that is not particularly effective in handling many outliers. This research aims to develop a control chart that is more resistant to outliers by using Minimum Regularized Covariance Determinant (MRCD). MRCD is a development of the MCD method, which is better at dealing with ‘fat data,” namely, situations in which the number of variables is greater than the number of observations. The performance of a robust Max-Half-Mchart control chart based on MRCD was evaluated using the average run length (ARL) against shifts in the process mean, process variance, and simultaneous shifts. A comparison was also made of the outlier detection accuracy between the robust Max-Half-Mchart based on MRCD and the standard Max-Half-Mchart. Simulation results demonstrated that the MRCD-based robust chart is most sensitive to simultaneous shifts in the mean and variance, significantly outperforming the conventional method in “de-masking” process deviations. The robust framework maintains higher accuracy and AUC levels even at extreme contamination stages of 30% to 40% outliers, where traditional charts typically fail. A practical application to cement quality data further substantiated these findings, as the robust chart successfully identified 14 out-of-control signals (comprising the mean, variability, and simultaneous shifts), whereas the conventional chart detected none. These results indicate that the MRCD-based Max-Half-Mchart offers a more reliable and responsive quality monitoring system for complex industrial datasets. Full article

The Guangdong–Hong Kong–Macao (GHM) region (especially the Greater Bay Area), a low-lying economic hub in southern China, faces complex particulate matter (PM_2.5) pollution dynamics under the combined influence of monsoonal systems and global warming. While long-term PM_2.5 reductions are documented, phase-specific trends remain obscured. Here, we analyze high-resolution ChinaHighPM2.5 dataset observations (2000–2023) using moving averages and piecewise regression to quantify abrupt shifts in interannual and seasonal PM_2.5 trends across the region. We identify 2014 and 2016 as critical breakpoints for annual PM_2.5 concentration (Mean_-5y-Year) and its linear acceleration rate (k_-5y-Year), respectively. Critical breakpoints delineate phases where declines persisted but decelerated. Prior to 2014, the PM_2.5 levels exhibited an upward trend (+0.203 µg·m⁻³·a⁻¹, p > 0.05), which reversed sharply post-2014 (−2.046 μg·m⁻³·a⁻¹, p < 0.01). Spatially, breakpoints clustered post-2014 for concentrations, while acceleration rate shifts reveal a latitudinal divergence near 23° N (23.873°~22.812° N); southern areas transitioned earlier (2010–2011) versus post-2014 in the north. Post-inflection declines are strongest toward the GBA urban core, with winter and autumn driving seasonal improvements (winter: steepest decline −2.646 μg·m⁻³·a⁻¹; autumn: largest trend reversal Δ−3.961 μg·m⁻³·a⁻¹), while improvement rates narrowed post-2016 (Δk = +0.527 µg·m⁻³·a⁻²). This study establishes that apparent regional PM_2.5 reductions mask significant spatiotemporal heterogeneity, underscoring the necessity of phase-specific analysis for effective pollution control in climatically vulnerable megaregions. Full article

Tularemia is a plague-like, potentially fatal zoonosis caused by the coccobacillus Francisella tularensis. It was discovered at the beginning of the last century in the United States and was soon recognized in Japan and in the former Soviet Union as the cause of clinical conditions that had been known for one and two centuries, respectively. More than 250 animal species are susceptible to infection, with rodents and lagomorphs serving as key reservoirs, and several vectors may transmit the disease, mainly ticks and mosquitoes. Humans are incidental hosts and are infected primarily by two F. tularensis subspecies, tularensis and holarctica: the former is more severe and is found almost exclusively in North America, whereas the latter is distributed throughout the Northern Hemisphere, mainly in Europe and Asia. Tularemia is highly infectious; therefore, diagnostic cultures should be handled in biosafety level 3 laboratories. Nevertheless, interhuman transmission is exceedingly rare. Although tularemia is relatively uncommon, it shows a re-emerging pattern at the global level, particularly in Europe. As with plague, mitigation may be more effectively achieved through a One Health approach. Neither approved vaccines nor therapeutic antibodies are currently available, whereas aminoglycoside, tetracycline, and quinolone antibiotics are effective. Owing to its high infectivity, its ease of transmission by inhalation, its clinical severity, with a prolonged and debilitating course, and its potential lethality, F. tularensis has long been considered a potential biological weapon, particularly if antibiotic-resistant strains were used. Although natural antibiotic resistance has not been described to date, research programs aimed at obtaining resistant strains have been conducted. It has been suggested that the disease was already present in the Middle East during the second millennium BC; should this hypothesis be confirmed by paleogenomic studies, plague and tularemia would have coexisted for more than three millennia, with plague masking the less severe tularemia. Many challenges related to tularemia are still unresolved. Full article

Video AI recognition is crucial for coal mine safety, but complex environments often yield low-quality images, hindering intelligent monitoring. Existing enhancement methods typically focus on image quality alone, lacking adaptability to specific tasks. Therefore, we propose Mine-DCE-YDT: a task-aware low-light image enhancement model that jointly optimizes enhancement with downstream object detection, ensuring enhanced images are both visually clearer and more conducive to accurate detection. Firstly, an improved Zero-DCE algorithm (Mine-DCE) is presented by introducing a Brightness-aware Mask Coordinate Attention (BMCA) module to improve illumination balance in the Value channel of the HSV image and a Multi-scale Detail Enhancement (MDE) module to reinforce textures and suppress noise. Then, Mine-DCE is co-modeled with YOLOv11n by training end-to-end via a joint loss fusing detection and enhancement quality losses to form Mine-DCE-YDT, which can enhance specific details containing image detection targets. Experimental results show that compared with Zero-DCE, Mine-DCE-YDT achieves reductions of 9.5% in NIQE and 35.5% in BRISQUE on the custom-constructed MineDataset and exhibits great enhancement performance on the public dataset LOL-V1. For the miner detection task in MineDataset, the integration of Mine-DCE-YDT with YOLOv11n achieves increases of 2.8% and 8.3% in mAP@0.5 and mAP@0.5:0.95, demonstrating its effectiveness in enhancing task-critical image features. Full article

Reliable process monitoring and quality evaluation for resistance spot welding (RSW) have become more important now than ever. An ultrasonic probe embedded into welding electrodes has enabled the acquisition of data about molten pool formation throughout welding, but automation of high-performance ultrasonic data analyses is still necessary to fully realize a monitoring system. This work proposes a two-stage deep learning (DL) approach for automated ultrasonic data analysis for RSW processing monitoring. The first stage conducts semantic segmentation on ultrasonic M-scan welding process signatures, yielding masks for identified molten pool and stack regions from which weld penetration measurements can be directly extracted, as well as expulsion occurrences throughout welding. From input images and segmentation outputs, the second stage directly estimates resultant weld nugget diameters using an additional neural network. Both stages leveraged architectures based on TransUNet, mixing elements of both convolutional neural networks (CNN) and vision transformers, and the effect of cross-attention for stack-up sheet thickness data fusion was investigated via an ablation study. Additionally, in the diameter estimation stage, the ablation study included alternative feature extraction architectures in the network and investigated the provision of M-scans to the model alongside segmentation masks. In both cases, cross-attention was determined to improve performance, and in the case of diameter estimation, providing M-scans as input was found to be beneficial in general. With cross-attention, the segmentation approach yielded a mean intersection over union (IoU) of 0.942 on molten pool, stack, and expulsion regions in the M-scans with 13.4 ms inference time. With cross-attention, diameter estimates yielded a mean absolute error of 0.432 mm with 4.3 ms inference time, representing a significant improvement over algorithmic approaches based on ultrasonic time of flight. Additionally, the approach attained >90% probability of detection (POD) at 0.830 mm below the acceptable diameter threshold and <10% probability of false alarm (PFA) at 0.828 mm above the threshold. These results demonstrate a novel production-ready application of DL in ultrasonic nondestructive evaluation (NDE) and pave the way for zero-defect RSW manufacturing. Full article

Few-shot object detection suffers from limited annotations, redundant background interference, insufficient feature interaction, and severe sample imbalance. Existing meta-learning-based methods extract class prototypes from support images but often fail to effectively suppress background noise or align channel-wise features between support and query branches. To address these issues, we propose a few-shot object detection method based on background reconstruction and multi-channel interactive feature fusion. First, a background reconstruction module is designed to suppress redundant background interference by applying random region masking to support set images, thereby generating robust class prototype features that are resistant to background noise. Second, a multi-channel interactive feature fusion module is designed, which leverages depthwise separable convolution to enable effective channel-wise feature interaction and information alignment between support class prototypes and query features, thereby enhancing cross-branch feature interaction and fusion. Finally, to address the uneven sample distribution and the foreground–background imbalance in few-shot scenarios, we proposed a category-aware weighted loss. By appropriately weighting the contributions of different object categories and background samples, the proposed loss encourages balanced optimization, resulting in faster convergence and improved detection performance. Experimental results demonstrate that the proposed method improves detection accuracy and generalization performance under few-shot settings. On the Pascal VOC dataset (Split1), the proposed method achieves 45.1%, 62.9%, and 67.1% under 1-shot, 3-shot, and 10-shot settings, respectively, outperforming the baseline; consistent improvements are also observed on the MS COCO dataset and the DIOR dataset. Full article

High-frequency non-intrusive load monitoring provides detailed harmonic information for appliances’ power disaggregation, and machine-learning approaches have demonstrated good performance in this task. However, these methods provide little transparency regarding the information structure of the aggregate signal. To address this, this paper models NILM as a coding-decoding process and applies information-theoretic measures to quantify uncertainty, recoverability, temporal contribution, and inter-appliance masking effects in aggregate signals. In the analyzed dataset, transfer entropy suggests negligible temporal gains, which is consistent with the observed effectiveness of pointwise models such as Random Forest. Moreover, conditional mutual information emphasizes the asymmetric masking relationships between appliances, with the laptop charger acting as a dominant interferer in the considered measurements. These findings are validated through a Random Forest regression model with minimum Redundancy Maximum Relevance feature selection. The results show that the mutual information between an appliance and the aggregate is a good predictor of disaggregation performance in the examined data, as appliances with high mutual information, such as hair dryer and electric water heater, achieve lower estimation errors, while others, such as iron, are difficult to recover despite stable distributions. This relationship is statistically supported by a strong negative monotonic correlation between normalized mutual information and the disaggregation error (Spearman $r_s=-0.81$, $p=0.015$). Hence, this work demonstrates how information-theoretic analysis can help characterize disaggregation difficulty prior to model training and assess the observability of appliances in high-frequency NILM. Full article

With the continuous growth in the scale of student physical health (SPH) monitoring data, annually sampled time-series records provide a valuable foundation for risk early warning. However, traditional models often fail to capture multi-year developmental trajectories of individuals, resulting in delayed intervention for students at potential health risk. This study aims to develop a Transformer-based Student Physical Anomaly Detection System (TSPADS), which is a dedicated intelligent software system to enable effective and timely anomaly detection in SPH data. The proposed TSPADS is built on the Transformer architecture and incorporates a novel masked anomaly-attention mechanism to learn implicit long-span dependencies in SPH data. A density-based clustering algorithm is then applied to distinguish anomalies and automatically generate hierarchical warning signals. Comprehensive experiments were conducted on a public multimodal movement and health dataset. The results demonstrate that TSPADS achieves high effectiveness and efficiency in both anomaly detection and classification tasks. The system shows strong potential to assist educational administrators and physical education teachers in providing timely, personalized health guidance, thereby addressing a critical gap in existing student health monitoring approaches. Full article

Synthetic dataset generation in Computer Vision, particularly for industrial applications, is still underexplored. Industrial defect segmentation, for instance, requires highly accurate labels, yet acquiring such data is costly and time-consuming. To address this challenge, we propose a novel diffusion-based pipeline for generating high-fidelity industrial datasets with minimal supervision. Our approach conditions the diffusion model on enriched bounding-box representations to produce precise segmentation masks, ensuring realistic and accurately localized defect synthesis. Compared to existing layout-conditioned generative methods, our approach improves defect consistency and spatial accuracy. We introduce two quantitative metrics to evaluate the effectiveness of our method and assess its impact on a downstream segmentation task trained on real and synthetic data. Our results demonstrate that diffusion-based synthesis can bridge the gap between artificial and real-world industrial data, fostering more reliable and cost-efficient segmentation models. Full article