Search Results (1,215)

Clinical photographs and patient metadata are complementary sources of information for automated skin lesion classification. In this study, a multimodal framework is proposed that combines Swin Transformer Tiny (Swin-T), Hyper-Feature Fusion (HFF), and Attention-Gated Clinical Metadata Fusion (AGCMF). HFF combines four levels of hierarchical representations from the Swin Transformer into a 256-dimensional visual embedding. A multilayer perceptron encodes 20 clinical metadata variables into a 64-dimensional representation. The features are concatenated, adaptively gated and projected for six-class classification. Five-fold patient-grouped cross-validation was performed on PAD-UFES-20, which consists of 2298 clinical photographs of 1373 patients. The proposed model obtained 83.55% accuracy without test-time augmentation (TTA), 83.29% with TTA, 82.92% weighted F1, 77.29% macro F1, 75.75% balanced accuracy, 93.12% macro-ROC-AUC, 94.59% weighted ROC-AUC, and Cohen’s kappa of 0.7702. In the ablation experiments, HFF increased the accuracy of the Swin-T model from 78.64% to 80.91%. Clinical feature concatenation increased it to 81.76%, and the complete HFF + AGCMF framework achieved 83.29%. These results indicate that hierarchical visual fusion and adaptive multimodal integration provide complementary benefits. The proposed framework achieved 3.89 percentage points higher accuracy than the image-plus-metadata baseline under the same TTA evaluation setting. A biopsy-exclusion sensitivity analysis also revealed that the performance of the model was only slightly affected by the removal of the diagnostic-verification variable, with the 19-feature model maintaining 82.91% accuracy, 82.53% weighted F1, and 94.21% weighted ROC-AUC. The most difficult category was still squamous cell carcinoma with 41.67% recall. The framework provided effective multimodal representation and improved class discrimination. However, independent external validation is required to assess its clinical generalizability. Full article

Mamba-based point cloud networks process serialized point tokens with state space layers and offer efficient classification, yet their performance is largely determined by the quality of local tokens produced before serialization. Existing tokenization compresses local patches via symmetric pooling, thereby discarding fine-grained geometric relationships and introducing aggregation bias under non-uniform sampling. To handle this, we propose a density-aware multi-level geometry enhancement method. A density-weighted token encoder estimates the local point density within each patch and adaptively calibrates point-wise contributions before aggregation, thus reducing the dominance of redundant dense samples. A multi-level local geometry branch extracts hierarchical coordinate-based neighborhood features directly from raw points and injects them into serialized tokens through residual fusion, compensating for geometric details lost during patch compression. Supervised contrastive learning is further adopted as an auxiliary regularizer to improve intra-class compactness and inter-class separability. Experiments on ScanObjectNN and ModelNet40 confirm the effectiveness of the approach: our method achieves 94.42 ± 0.09%, 92.11 ± 0.13%, and 87.89 ± 0.11% on the OBJ_BG, OBJ_ONLY, and PB_T50_RS variants, respectively, while ModelNet40 accuracy reaches 93.04 ± 0.12%. These results, obtained with only 12.57 M parameters, indicate a favorable accuracy–efficiency trade-off on the evaluated benchmarks. Full article

To address difficulties in traffic sign detection under low-light environments, this paper proposes RiDW-YOLO, an improved detection algorithm based on YOLOv11n. The Retinexformer network is embedded as a trainable front-end module at the first layer of the YOLOv11n backbone, performing online image enhancement during forward propagation without offline preprocessing. Its weights are updated end-to-end with the subsequent detection sub-network, improving brightness and contrast while suppressing noise, thereby strengthening feature extraction for traffic sign targets. An iterative attentional feature fusion (iAFF) block is integrated into the feature-fusion architecture of YOLOv11n, enabling adaptive weighted multi-level feature aggregation and enhancing feature representation. DySample, a dynamic up-sampling operator, replaces conventional interpolation methods by learning offset coordinates to better preserve fine-grained feature information. Wise-IoU (WIoU) replaces the original Complete-Intersection over Union (CIoU) loss function, leveraging dynamically adjusted gradient weights to suppress low-quality samples interference and boost bounding-box localization performance. Experimental results demonstrate that compared with the YOLOv11n baseline model, the improved algorithm achieves an increase of 15.2 percentage points in precision, 6.2 percentage points in recall, and 11.0 percentage points in mAP@50. Overall, this work provides a feasible solution for traffic-sign detection under low-light conditions and emphasizes the importance of balanced module design and frank discussion of current limitations. Full article

Background: Differentiating alcohol-related liver disease (ARLD) from chronic liver injury caused by other etiologies remains a clinically relevant challenge in cross-sectional imaging. In particular, alcoholic hepatitis and alcoholic cirrhosis may overlap morphologically with HBV-related cirrhosis on CT imaging. To develop and assess an attention-residual hybrid Convolutional Neural Network (CNN) for multiclass CT image classification of alcoholic hepatitis, alcoholic cirrhosis, HBV-related cirrhosis, and living liver donors. Methods: A four-class liver image dataset comprising 5760 CT images from 144 individuals (36 per group) was constructed using images from alcoholic hepatitis, alcoholic cirrhosis, HBV-related cirrhosis, and living liver donors. The dataset was partitioned into training, validation, and test sets at the patient level. Five pretrained CNN architectures, including DenseNet121, ResNet50, MobileNetV3-Large, EfficientNetB0, and ConvNeXt-Tiny, were first fine-tuned and comparatively evaluated. Based on F1-score ranking, DenseNet121 and ConvNeXt-Tiny were selected as the two backbone networks for the proposed hybrid model. The final architecture integrated Attention Pooling, Feature-wise Linear Modulation (FiLM), Multi-head Attention, Gated Linear Units, Residual Connections, and layer normalization to improve feature fusion and contextual representation. Results: The proposed model demonstrated the best overall performance among all evaluated architectures on the test set. It achieved an accuracy of 99.55%, a weighted F1-score of 99.55%, an MCC of 0.9941, and a Cohen’s kappa coefficient of 0.9940. The model also achieved ROC-AUC and PR-AUC values of 100.00% and 99.99%, respectively, together with an NPV of 99.85%. The proposed model’s performance was also balanced across classes. For Alcoholic Cirrhosis, precision, recall, and F1-score were all 99.64%. For Alcoholic Hepatitis, the corresponding values were 100.00%, 98.93%, and 99.46%, respectively, while HBV-related Cirrhosis achieved 99.29% precision, 99.64% recall, and 99.47% F1-score. Living Liver Donors achieved 99.29% precision, 100.00% recall, and 99.64% F1-score. Conclusions: The findings of this exploratory study suggest that routine CT images may contain image-based differences potentially relevant to etiology-oriented classification of diffuse liver disease. Beyond distinguishing ARLD from HBV-related cirrhosis and images from living liver donors, the model also captured image-based differences between major ARLD subgroups, including alcoholic hepatitis and alcoholic cirrhosis. These findings support further investigation of CT-derived image-based differences in larger independent and multicenter datasets. Full article

To improve prompt utilization and boundary perception in SAM-based interactive medical image segmentation, this paper proposes ASAM2-UNet, a U-shaped visual foundation model built upon SAM2-UNet. Unlike methods that treat prompts as auxiliary spatial inputs, ASAM2-UNet explicitly incorporates prompt-derived priors into feature reasoning through an Interactive Prompt-Guided Focal Attention module. Specifically, a user-provided spatial prompt is converted into an explicit attention prior that modulates focal self-attention, allowing the network to emphasize target-relevant regions while retaining efficient local–global contextual modeling. In addition, a Contextual Semantic Information Complement module integrates multi-scale decoder features with uncertainty-aware and structure-aware cues derived from the initial prediction to refine ambiguous lesion boundaries. By integrating cross-layer spatial attention, local–global spatial fusion, and semantic-aware feature aggregation, the proposed module enhances the discrimination of ambiguous edges and complex foreground–background regions. Extensive experiments on five public polyp segmentation datasets demonstrate the effectiveness of the proposed method. Full article

In recent years, the use of UAVs has become increasingly widespread, and the public safety risks posed by unauthorized UAV flights have become increasingly prominent, creating an urgent need for effective detection and identification of UAV targets. However, such targets are small in size, have low contrast, and exhibit an extremely low signal-to-noise ratio; conventional detection methods generally suffer from insufficient feature discrimination, missed detections, and false alarms in complex backgrounds. To address these challenges, this paper proposes a Frequency-Guided Cross-scale Refinement Network (FGCR-Net). Based on an encoder-decoder architecture, this network achieves end-to-end collaborative optimization through cross-layer feature fusion, side-channel prediction refinement, and frequency-domain background suppression. First, a multi-path selective cross-layer fusion module (SCFM) is designed. This module employs coordinated modeling via both channel and spatial paths, supplemented by adaptive weighting with learnable coefficients, to perform differentiated selective fusion of the encoder’s fine-grained features and the decoder’s semantic features, thereby bridging the semantic gap at jump connections; Second, we designed a Cross-Scale Adaptive Fusion Enhancement Attention Module (CAFEM), which cascades multi-receptive-field hollow convolutions, strip pooling, and a bidirectional semantic guidance mechanism to perform cross-scale refinement on the side outputs of each decoder layer, thereby alleviating the issues of blurred boundaries and false alarms caused by inconsistent quality of multi-scale prediction maps and insufficient cross-layer consistency; finally, we design a Frequency-Guided Semantic Enhancement Module (FGSEM), which uses the Fast Fourier Transform (FFT) to decouple encoder features into the frequency domain. By leveraging low-frequency energy to predict the background confidence map and applying spatially selective suppression to high-frequency components, this module distinguishes, from a frequency-domain perspective, the high-frequency responses of complex backgrounds and targets that are highly similar in the spatial domain. Experiments on MSDS-UAV, a self-built multi-scenario UAV dataset for small targets, demonstrate that our method consistently outperforms existing state-of-the-art methods across multiple performance metrics, with Pixel Accuracy, Mean Intersection over Union, and Probability of Detection reaching 92.76%, 70.91%, and 92.69%, respectively; Compared to the baseline model, these three metrics improved by 1.90, 3.20, and 3.76 percentage points, respectively, fully validating the effectiveness and superiority of the proposed method. Full article

The increasing interconnection of smart-grid operational technology, industrial-edge services, and utility information systems creates a critical need for resilient and continuously monitored industrial cyber-physical communication. Although quantum key distribution (QKD) can strengthen session-key establishment for advanced metering infrastructure, distributed energy resources, substation automation, supervisory control, and utility-core services, practical QKD deployments remain vulnerable to implementation-level side-channel attacks that can compromise the cryptographic protection layer without directly targeting conventional network packets. This paper presents a QKD-secured industrial smart-grid cyber-physical system framework for simulating and detecting adaptive side-channel attacks. The proposed 36-node industrial communication architecture integrates AMI devices, DER controllers, PMU and substation automation components, industrial-edge gateways, QKD modules, key-management services, SCADA and utility-core servers, security-operation-center components, and adversarial access points. A 100,000-record cyber-quantum dataset is generated across 12 operating conditions comprising normal communication and 11 adaptive QKD side-channel attacks: detector blinding, time shift, wavelength switching, Trojan-horse probing, photon-number splitting, decoy-state spoofing, RNG bias, calibration manipulation, local-oscillator manipulation, synchronization spoofing, and combined adaptive quantum hacking. Each scenario introduces coupled primary and secondary perturbations across optical, detector, timing, synchronization, randomness, calibration, photon-statistical, leakage, key-generation, encryption, and industrial-network-performance features. To support intelligent industrial security monitoring, the proposed Quantum-aware Mamba–Kolmogorov–Arnold Network (Q-MambaKAN) organizes device, network, QKD, side-channel, encryption, and risk evidence into an ordered cyber-quantum representation processed through selective state-space learning, side-channel attention, nonlinear KAN mapping, adaptive fusion, and multi-task prediction heads. Results show that the QBER increases from 0.071 during normal operation to 0.426 under combined adaptive quantum hacking, while encryption success decreases from 98.1% to 0%. Q-MambaKAN achieves a 99.48% binary detection accuracy, a 99.70% binary F1-score, a 97.60% multiclass macro-F1, and a risk RMSE of 0.021. Full article

Remote Sensing Image Super-Resolution (RSISR) is a core task in geospatial image analysis. Convolutional neural networks (CNNs) have achieved significant breakthroughs in RSISR tasks by extracting local features. However, CNN-based methods struggle to capture long-range dependencies, thereby limiting SR performance. Recently, Transformer-based methods have demonstrated remarkable performance in capturing global information. Nevertheless, they remain inadequate for exploring high-frequency details and local features. To overcome these limitations, this work introduces a novel dual-path collaborative architecture, named DCTNet, which combines Transformer-based global modeling with convolution-driven local feature extraction. DCTNet is a hybrid network composed of a CNN-Transformer Residual Hybrid Group (CTHG). This group consists of two core components: the Dual-domain Fusion Window Attention Block (DFWAB) and the Stepwise Dilated Convolution (SDC). Specifically, the DFWAB incorporates channel and frequency attention mechanisms following the standard Transformer block to recover high-frequency details. Furthermore, by integrating stepwise dilated convolutions into the conventional Transformer architecture, the CTHG effectively captures both multi-scale local and global features. Additionally, we employ dense connections among the DFWAB modules to facilitate feature reuse across layers. Experimental results on the AID and UCMerced datasets demonstrate that DCTNet achieves competitive reconstruction performance across different scale factors, with statistically significant improvements observed in specific settings. Full article

Object detection in visible nearshore surveillance imagery is of great importance for maritime safety, intelligent coastal monitoring, and water rescue applications. Nevertheless, reliable detection remains difficult because nearshore scenes often contain numerous small targets, cluttered wave patterns, shoreline textures, and substantial illumination variations. Moreover, existing public datasets mainly emphasize vessel detection and provide limited nearshore object categories. To solve these limitations, this study presents a large-scale visible nearshore dataset containing 20,934 images annotated with seven categories: pedestrian, sailor, swimmer, ship, boat, flotage, and seamark. The dataset is designed to support comprehensive evaluation and fair comparison of detection algorithms in complex nearshore environments. Based on the proposed benchmark, we conduct extensive evaluations of multiple mainstream object detectors and further develop a detection framework termed VN-DETR. The proposed model enhances both feature extraction and multi-scale feature fusion for nearshore scenarios. Specifically, a kernel selective attention based on WTConv (WKSA) module is designed to enlarge the receptive field and exploit contextual information in visible images, enabling more accurate object classification. In addition, a cross-layer feature selection and fusion (CFSF) module is introduced to perform feature matching, selection, and fusion across adjacent layers, enhancing the discriminability between foreground objects and complex nearshore backgrounds. This design effectively improves robustness against background noise such as wave reflections and shoreline textures. Extensive experiments on the constructed dataset demonstrate that VN-DETR consistently outperforms representative baseline methods and achieves superior detection performance, particularly for challenging small object categories. Full article

Traditional indoor localization using received signal strength (RSS) fingerprinting commonly uses measured RSS values as location-related inputs. This paper investigates a paired light–dark representation of an RSS fingerprint image and a cross-modality multi-scale fusion model for small-scale indoor localization. Specifically, each RSS fingerprint image is inverted in pixel intensity, and the original and inverted images are retained as synchronized representations of the same RSS measurement rather than as independent physical modalities or separate augmented samples. The paired images are processed by two parallel convolutional neural network (CNN) paths combined with a cross-modality fusion (CMF) module, which extracts and fuses multi-scale feature representations. The fused features are fed into fully connected layers to predict the coordinate vector. Experimental comparisons with traditional and advanced methods on four public datasets show positioning errors comparable to the strongest advanced benchmark and lower errors than the considered traditional baselines. The proposed model uses approximately 30.8% fewer parameters than the advanced benchmark. Full article

Accurate one-week-ahead building electricity demand forecasting is essential for building energy management, yet representing future building operational characteristics remains challenging because such information is generally unavailable in advance. This study investigates the effectiveness of representing building operational characteristics using cluster labels derived from daily electricity consumption patterns for medium-term electricity demand forecasting. Cluster labels obtained by k-means clustering were incorporated as operational-state features into a Temporal Fusion Transformer (TFT) together with historical electricity consumption, meteorological variables, and calendar information. Forecasting performance was evaluated for a training facility and three university buildings using walk-forward validation under different feature reference periods. Under an idealized information condition in which meteorological variables and cluster labels corresponding to the forecasting period were provided as known future inputs, this forecasting pattern achieved the highest accuracy for all investigated buildings. Under the same idealized condition, variable importance analysis indicated that the cluster label exhibited the highest importance among the known future inputs, exceeding that of calendar variables and most meteorological variables. In addition, the TFT outperformed Long Short-Term Memory (LSTM) and Multi-Layer Perceptron (MLP) models. These findings indicate the potential value of the proposed operational-state representation for improving one-week-ahead building electricity demand forecasting and provide interpretable insights into the contribution of operational-state features. Full article

Geometric deviations in unsupported overhang features pose one of the most persistent quality challenges in Laser Powder Bed Fusion (LPBF), where even small deviations from the intended geometry can undermine part functionality and reliability. This study presents a geometrically informed conditional Generative Adversarial Network (cGAN), implemented through the Pix2Pix framework, to predict layer-wise geometric deviations in LPBF-printed parts with overhang geometries, using paired two-dimensional Computer-Aided Design (2D CAD) slices and corresponding X-ray Computed Tomography (XCT)-derived ground truth slices. The study investigates how geometric information can be encoded within the conditional input of the Pix2Pix framework to more effectively guide deviation prediction. A total of 18 models were trained and evaluated across multiple overhang geometry groups and batch size configurations, assessed through a combination of perceptual, structural, and boundary-focused metrics, namely Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Learned Perceptual Image Patch Similarity (LPIPS), Fréchet Inception Distance (FID), and Edge Intersection over Union (Edge IoU). The results demonstrated that color-coded inputs consistently improved prediction fidelity, perceptual similarity, and edge alignment relative to their non-color-coded counterparts. Furthermore, a model trained on a balanced multi-geometry dataset showed improved prediction performance on withheld 30° and 60° overhang configurations within the benchmark geometry family. The proposed framework offers a data-driven, design-stage tool for anticipating geometry-dependent deviations in LPBF overhang structures, supporting design for additive manufacturing. Full article

Accurate and efficient monitoring of tea plant growth parameters via remote sensing is essential for precision plantation management. However, spectral indices relying solely on reflectance often exhibit limited sensitivity in capturing complex tea canopy characteristics. This study developed a data-driven framework integrating spectral reflectance, frequency-domain harmonic components, and spatial texture features to construct tri-feature fusion indices (TFIs) and establish machine learning and deep learning models for tea growth monitoring. Ten-band multispectral imagery was acquired using a UAV alongside synchronous field measurements of leaf and plant biomass and nitrogen accumulation. TFIs were constructed through exhaustive feature combinations and optimized via a data-driven search strategy. Subsequently, random forest (RF), multilayer perceptron (MLP), convolutional neural network (CNN), and transformer models were evaluated using a leave-one-site-out cross-validation (LOSO-CV) strategy. The selected TFIs showed strong associations with tea growth parameters within the investigated dataset, with R² values up to 0.63 and 0.62 for leaf dry matter and leaf nitrogen accumulation, respectively. Models incorporating selected TFIs achieved cross-validated R² values of 0.56 for leaf dry matter (MLP), 0.59 for plant dry matter (MLP), 0.73 for leaf nitrogen accumulation (MLP), and 0.68 for plant nitrogen accumulation (CNN). These models exhibited competitive predictive performance comparable to RF, although no statistically significant differences in mean absolute error were observed under site-held-out evaluation. Furthermore, model-derived spatial maps provided insights into fine-scale spatial heterogeneity and potential interannual variations in tea growth parameters across representative plantations from 2024 to 2025. Overall, this study provides a UAV-based framework for tea growth parameter estimation by integrating multi-domain information without requiring additional environmental observations. Full article

High-order phase defocus occurs in ISAR imaging of maneuvering targets. Traditional methods are limited by motion model mismatches, whereas existing deep imaging relies on complex-valued convolutions or self-attention mechanisms, whose excessive computational complexity hinders real-time processing. To address these issues, this paper proposes an Azimuth-Frequency Conditioned Mamba U-Net (AFC-Mamba U-Net) for ISAR image refocusing. The method treats the real and imaginary parts of the defocused complex image as dual-channel real-valued inputs, integrating linear-complexity state-space modeling with cross-layer feature fusion to significantly reduce computational and memory overhead while maintaining high refocusing quality. Specifically, we design an Azimuth-Frequency Conditional Mamba block (AFCMambaBlock), which performs bidirectional selective state-space scanning along the azimuth axis to capture long-range defocus dependencies with linear time complexity, and introduces azimuth-frequency conditional gating to incorporate frequency-domain phase error information into spatial features. A phase-state bottleneck is constructed to aggregate multi-scale enhanced states via a cross-scale state bank, forming globally phase-error-aware representations. Furthermore, we propose a state-guided skip connection mechanism that dynamically gates encoder skip features using bottleneck states, effectively suppressing the propagation of defocus artifacts while reducing redundant computations in the decoder. Experiments conducted on a simulated dataset containing eight types of spaceborne targets and real-world Yak-42 aircraft data demonstrate that the proposed method reduces MSE by 35.8% and NMSE by 41.5% compared to CNN-based U-Net, while achieving an SSIM of 0.9901. Compared to the state-of-the-art complex-domain network CVPHD, the proposed method achieves a significantly better efficiency–accuracy trade-off: it reduces model parameters and computational complexity by approximately two orders of magnitude, improves inference speed by more than fivefold, and cuts GPU memory usage by about 96%, while maintaining competitive structural consistency (SSIM > 0.99). Although CVPHD retains better MSE and NMSE on simulated data, the proposed method provides a lightweight and feasible solution for real-time and embedded ISAR refocusing. Full article

Face detection serves as the prerequisite and foundation for individual sheep and goat identification. Considering the practical detection process, the facial region of sheep and goats shares similarities with the torso texture features, which could be easily interfered with by the background and other factors, along with the presence of varying face scales. RetinaFace was selected as the framework, with GhostNet used as the backbone for feature extraction, thereby significantly reducing model complexity and computational load. A new feature extraction unit—SRM module—was designed through the integration of the Multi-Source Domain Adaptation (MSDA) attention mechanism and Reparameterized Non-local Cross-scale Feature Exchange Layer 4 (RepNCSPELAN4) to mitigate the impact of environmental factors and enhance the fine-grained feature extraction of the sheep and goat’s face. A new feature fusion network—A_EFPN—was designed through the com-bination of Efficient Reparameterized Generalized-FPN (RepGFPN) and Alterable Kernel Convolution (AKConv), thereby enhancing feature interaction and improving the model’s multi-scale detection ability. An optimized multi-task joint loss function was employed to retain three facial key points: the left eye, right eye, and nose, thereby reducing the number of model regression parameters. Experimental results demonstrate that the proposed method achieved 96.15% Precision, 97.06% Recall, 96.60% F1 Score, 98.31% AP, 2.40% NME, 2.36 G FLOPs, 3.63 M Params, and a weight of 14.75 MB. It was demonstrated to be applicable to the rapid and accurate detection of sheep and goat faces in the actual breeding environment, providing the foundation for subsequent sheep and goat face recognition. Full article