Search Results (2,840)

Accurate segmentation of fruit tree leaf diseases is critical for yield protection and precision crop management, yet it is challenging due to complex field conditions, irregular leaf morphology, and diverse lesion patterns. To address these issues, Disease-Seg, a lightweight real-time segmentation framework, is proposed. It integrates CNN and Transformer with a parallel fusion architecture to capture local texture and global semantic context. The Extended Feature Module (EFM) enlarges the receptive field while retaining fine details. A Deep Multi-scale Attention mechanism (DM-Attention) allocates channel weights across scales to reduce redundancy, and a Feature-weighted Fusion Module (FWFM) optimizes integration of heterogeneous feature maps, enhancing multi-scale representation. Experiments show that Disease-Seg achieves 90.32% mIoU and 99.52% accuracy, outperforming representative CNN, Transformer, and hybrid-based methods. Compared with HRNetV2, it improves mIoU by 6.87% and FPS by 31, while using only 4.78 M parameters. It maintains 69 FPS on 512 × 512 crops and requires approximately 49 ms per image on edge devices, demonstrating strong deployment feasibility. On two grape leaf diseases from the PlantVillage dataset, it achieves 91.19% mIoU, confirming robust generalization. These results indicate that Disease-Seg provides an accurate, efficient, and practical solution for fruit leaf disease segmentation, enabling real-time monitoring and smart agriculture applications. Full article

Short-term load forecasting (STLF) underpins the efficient and secure operation of power systems. This study develops and evaluates a hybrid architecture that couples deep neural networks (DNNs) with support vector regression (SVR) for national-scale day-ahead STLF using Croatian load data from 2006 to 2022. The approach employs an hour-specific framework of 24 hybrid models: each DNN learns a compact nonlinear representation for a given hour, while an SVR trained on the penultimate layer activations performs the final regression. Gradient-boosting-based feature selection yields compact, informative inputs shared across all model variants. To overcome limitations of historical local measurements, the framework integrates global numerical weather prediction data from the TIGGE archive with load and local meteorological observations in an operationally realistic setup. In the held-out test year 2022, the proposed hybrid consistently reduced forecasting error relative to standalone DNN-, LSTM- and Transformer-based baselines, while preserving a reproducible pipeline. Beyond using SVR as an alternative output layer, the contributions are as follows: addressing a 17-year STLF task, proposing an hour-specific hybrid DNN–SVR framework, providing a systematic comparison with deep learning baselines under a unified protocol, and integrating global weather forecasts into a practical day-ahead STLF solution for a real power system. Full article

The extraction and matching of interest points are fundamental to many geometric computer vision tasks. Traditionally, matching is performed by assigning descriptors to interest points and identifying correspondences based on descriptor similarity. This work introduces a technique whereby interest points are inherently associated during detection, eliminating the need for computing, storing, transmitting, or matching descriptors. Although the matching accuracy is marginally lower than that of conventional approaches, our method completely eliminates the need for descriptors, leading to a drastic reduction in memory usage for localization systems. We assess its effectiveness by comparing it against both classical handcrafted methods and modern learned approaches. Full article

Defect inspection of Printed Circuit Board (PCB) is essential for maintaining the safety and reliability of electronic products. With the continuous trend toward smaller components and higher integration levels, identifying tiny imperfections on densely packed PCB structures has become increasingly difficult and remains a major challenge for current inspection systems. To tackle this problem, this study proposes the Multi-scale Edge-Aware Enhanced Detection Transformer (MEE-DETR), a deep learning-based object detection method. Building upon the RT-DETR framework, which is grounded in Transformer-based machine learning, the proposed approach systematically introduces enhancements at three levels: backbone feature extraction, feature interaction, and multi-scale feature fusion. First, the proposed Edge-Strengthened Backbone Network (ESBN) constructs multi-scale edge extraction and semantic fusion pathways, effectively strengthening the structural representation of shallow defect edges. Second, the Entanglement Transformer Block (ETB), synergistically integrates frequency self-attention, spatial self-attention, and a frequency–spatial entangled feed-forward network, enabling deep cross-domain information interaction and consistent feature representation. Finally, the proposed Adaptive Enhancement Feature Pyramid Network (AEFPN), incorporating the Adaptive Cross-scale Fusion Module (ACFM) for cross-scale adaptive weighting and the Enhanced Feature Extraction C3 Module (EFEC3) for local nonlinear enhancement, substantially improves detail preservation and semantic balance during feature fusion. Experiments conducted on the PKU-Market-PCB dataset reveal that MEE-DETR delivers notable performance gains. Specifically, Precision, Recall, and mAP_50–95 improve by 2.5%, 9.4%, and 4.2%, respectively. In addition, the model’s parameter size is reduced by 40.7%. These results collectively indicate that MEE-DETR achieves excellent detection performance with a lightweight network architecture. Full article

With the rapid advancement of smart grids driven by renewable energy integration and the extensive deployment of supervisory control and data acquisition (SCADA) and phasor measurement units (PMUs), addressing the escalating alarm flooding via intelligent analysis of large-scale alarm data is pivotal to safeguarding the safe and stable operation of power grids. To tackle these challenges, this study introduces a pioneering alarm optimization framework based on symmetry-driven crowdsourced active learning and interpretable deep reinforcement learning (DRL). Firstly, an anomaly alarm annotation method integrating differentiated crowdsourcing and active learning is proposed to mitigate the inherent asymmetry in data distribution. Secondly, a symmetrically structured DRL-based hierarchical attention deep Q-network is designed with a dual-path encoder to balance the processing of multi-scale alarm features. Finally, a SHAP-driven interpretability framework is established, providing global and local attribution to enhance decision transparency. Experimental results on a real-world power alarm dataset demonstrate that the proposed method achieves a Fleiss’ Kappa of 0.82 in annotation consistency and an F1-Score of 0.95 in detection performance, significantly outperforming state-of-the-art baselines. Additionally, the false positive rate is reduced to 0.04, verifying the framework’s effectiveness in suppressing alarm flooding while maintaining high recall. Full article

Collision avoidance between unmanned aerial vehicles (UAVs) and non-cooperative targets (e.g., off-nominal operations or birds) presents significant challenges in urban air mobility (UAM). This difficulty arises due to the highly dynamic and unpredictable flight intentions of these targets. Traditional collision-avoidance methods primarily focus on cooperative targets or non-cooperative ones with fixed behavior, rendering them ineffective when dealing with highly unpredictable flight patterns. To address this, we introduce a deep reinforcement learning-based collision-avoidance approach leveraging global and local intent prediction. Specifically, we propose a Global and Local Perception Prediction Module (GLPPM) that combines a state-space-based global intent association mechanism with a local feature extraction module, enabling accurate prediction of short- and long-term flight intents. Additionally, we propose a Fusion Sector Flight Control Module (FSFCM) that is trained with a Dueling Double Deep Q-Network (D3QN). The module integrates both predicted future and current intents into the state space and employs a specifically designed reward function, thereby ensuring safe UAV operations. Experimental results demonstrate that the proposed method significantly improves mission success rates in high-density environments, with up to 80 non-cooperative targets per square kilometer. In 1000 flight tests, the mission success rate is 15.2 percentage points higher than that of the baseline D3QN. Furthermore, the approach retains an 88.1% success rate even under extreme target densities of 120 targets per square kilometer. Finally, interpretability analysis via Deep SHAP further verifies the decision-making rationality of the algorithm. Full article

In complex wireless communication environments, automatic modulation classification (AMC) faces two critical challenges: the lack of robustness under low-signal-to-noise ratio (SNR) conditions and the inefficiency of integrating multi-scale feature representations. To address these issues, this paper proposes CAIC-Net, a robust modulation classification network that integrates a dynamic cross-attention mechanism with a cross-SNR contrastive learning strategy. CAIC-Net employs a dual-stream feature extractor composed of ConvLSTM2D and Transformer blocks to capture local temporal dependencies and global contextual relationships, respectively. To enhance fusion effectiveness, we design a Dynamic Cross-Attention Unit (CAU) that enables deep bidirectional interaction between the two branches while incorporating an SNR-aware mechanism to adaptively adjust the fusion strategy under varying channel conditions. In addition, a Cross-SNR Contrastive Learning (CSCL) module is introduced as an auxiliary task, where positive and negative sample pairs are constructed across different SNR levels and optimized using InfoNCE loss. This design significantly strengthens the intrinsic noise-invariant properties of the learned representations. Extensive experiments conducted on two standard datasets demonstrate that CAIC-Net achieves competitive classification performance at moderate-to-high SNRs and exhibits clear advantages in extremely low-SNR scenarios, validating the effectiveness and strong generalization capability of the proposed approach. Full article

Deepfake detection in video data typically relies on mining deep embedded representations across multiple modalities to obtain discriminative fused features and thereby improve detection accuracy. However, existing approaches predominantly focus on how to exploit complementary information across modalities to ensure effective fusion, while often overlooking the impact of noise and interference present in the data. For instance, issues such as small objects, blurring, and occlusions in the visual modality can disrupt the semantic consistency of the fused features. To address this, we propose a Multimodality Semantic Consistency Fusion model for video forgery detection. The model introduces a semantic consistency gating mechanism to enhance the embedding of semantically aligned information across modalities, thereby improving the discriminability of the fused representations. Furthermore, we incorporate an event-level weakly supervised loss to strengthen the global semantic discrimination of the video data. Extensive experiments on standard video forgery detection benchmarks demonstrate the effectiveness of the proposed method, achieving superior performance in both forgery event detection and localization compared to state-of-the-art approaches. Full article

Coalbed methane, an abundant clean energy resource in China, is gaining significant attention. Electric submersible progressive cavity pumps, ideal for downhole extraction with high solids content, are vital in coalbed methane operations. Current fault diagnosis research for these pumps mainly relies on machine learning algorithms to identify fault features, but complex working conditions and imbalanced sample distributions challenge these models’ ability to perceive multi-scale and multi-dimensional features. To enhance the model’s perception of deep abnormal data in complex multi-case industrial datasets, this study proposes a deep learning model based on a multi-scale extraction and residual module convolutional neural network. Innovatively, a cross-attention module using global autocorrelation and local cross-correlation is introduced to constrain the multi-scale feature extraction process, making the model better suited to specific and differentiated data environments. Post feature extraction, the model employs Borderline-SMOTE to augment minority class samples and uses Tomek Links for noise removal. These enhancements improve the comprehensive perception of fault types with significant differences in period, amplitude, and dimension, as well as the learning capability for rare faults. Based on field-collected fault data and using enhanced and cleaned features for classifier training, tests on a real industrial dataset show the proposed model achieves an F1 Measure of 90.7%—an improvement of 13.38% over the unimproved model and 9.15–31.64% over other common fault diagnosis models. Experimental results confirm the method’s effectiveness in adapting to extremely imbalanced sample distributions and complex, variable field data characteristics. Full article

Magnetic resonance imaging (MRI) super-resolution (SR) enables high-resolution reconstruction from low-resolution acquisitions, reducing scan time and easing hardware demands. However, most deep learning-based SR models are large and computationally heavy, limiting deployment in clinical workstations, real-time pipelines, and resource-restricted platforms such as low-field and portable MRI. We introduce CHARMS, a lightweight convolutional–Transformer hybrid with attention regularization optimized for MRI SR. CHARMS employs a Reverse Residual Attention Fusion backbone for hierarchical local feature extraction, Pixel–Channel and Enhanced Spatial Attention for fine-grained feature calibration, and a Multi-Depthwise Dilated Transformer Attention block for efficient long-range dependency modeling. Novel attention regularization suppresses redundant activations, stabilizes training, and enhances generalization across contrasts and field strengths. Across IXI, Human Connectome Project Young Adult, and paired 3T/7T datasets, CHARMS (~1.9M parameters; ~30 GFLOPs for 256 × 256) surpasses leading lightweight and hybrid baselines (EDSR, PAN, W2AMSN-S, and FMEN) by 0.1–0.6 dB PSNR and up to 1% SSIM at ×2/×4 upscaling, while reducing inference time ~40%. Cross-field fine-tuning yields 7T-like reconstructions from 3T inputs with ~6 dB PSNR and 0.12 SSIM gains over native 3T. With near-real-time performance (~11 ms/slice, ~1.6–1.9 s per 3D volume on RTX 4090), CHARMS offers a compelling fidelity–efficiency balance for clinical workflows, accelerated protocols, and portable MRI. Full article

Accurate indoor positioning and navigation remain significant challenges, with audio sensor-based sound source localization emerging as a promising sensing modality. Conventional methods, often reliant on multi-channel processing or time-delay estimation techniques such as Generalized Cross-Correlation, encounter difficulties regarding computational complexity, hardware synchronization, and reverberant environments where time difference in arrival cues are masked. While machine learning approaches have shown potential, their performance depends heavily on the discriminative power of input features. This paper proposes a novel feature extraction method named Short-Time Homomorphic Deconvolution, which transforms multi-channel audio signals into a 2D Time × Time-of-Flight representation. Unlike prior 1D methods, this feature effectively captures the temporal evolution and stability of time-of-flight differences between microphone pairs, offering a rich and robust input for deep learning models. We validate this feature using a lightweight Convolutional Neural Network integrated with a dual-stage channel attention mechanism, designed to prioritize reliable spatial cues. The system was trained on a large-scale dataset generated via simulations and rigorously tested using real-world data acquired in an ISO-certified anechoic chamber. Experimental results demonstrate that the proposed model achieves precise Direction of Arrival estimation with a Mean Absolute Error of 1.99 degrees in real-world scenarios. Notably, the system exhibits remarkable consistency between simulation and physical experiments, proving its effectiveness for robust indoor navigation and positioning systems. Full article

The energy transition, driven by the global shift toward renewable and electrification, necessitates accurate and efficient prediction of electrical load profiles to quantify energy consumption. Therefore, the systematic literature review (SLR), followed by PRISMA guidelines, synthesizes hybrid architectures for sequential electrical load profiles, aiming to span statistical techniques, machine learning (ML), and deep learning (DL) strategies for optimizing performance and practical viability. The findings reveal a dominant trend towards complex hybrid models leveraging the combined strengths of DL architectures such as long short-term memory (LSTM) and optimization algorithms such as genetic algorithm and Particle Swarm Optimization (PSO) to capture non-linear relationships. Thus, hybrid models achieve superior performance by synergistically integrating components such as Convolutional Neural Network (CNN) for feature extraction and LSTMs for temporal modeling with feature selection algorithms, which collectively capture local trends, cross-correlations, and long-term dependencies in the data. A crucial challenge identified is the lack of an established framework to manage adaptable output lengths in dynamic neural network forecasting. Addressing this, we propose the first explicit idea of decoupling output length predictions from the core signal prediction task. A key finding is that while models, particularly optimization-tuned hybrid architectures, have demonstrated quantitative superiority over conventional shallow methods, their performance assessment relies heavily on statistical measures like Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). However, for comprehensive performance assessment, there is a crucial need for developing tailored, application-based metrics that integrate system economics and major planning aspects to ensure reliable domain-specific validation. Full article

The advancement of hyperspectral image super-resolution (HSI-SR) has been significantly propelled by deep learning techniques. However, current methods predominantly rely on 2D or 3D convolutional networks, which are inherently local and thus limited in modeling long-range spectral–depth interactions. This work introduces a novel network architecture designed to address this gap through recursive deep feature learning. Our model initiates with 3D convolutions to extract preliminary spectral–spatial features, which are progressively refined via densely connected grouped convolutions. A core innovation is a recursively formulated generalized self-attention mechanism, which captures long-range dependencies across the spectral dimension with linear complexity. To reconstruct fine spatial details across multiple scales, a progressive upsampling strategy is further incorporated. Evaluations on several public benchmarks demonstrate that the proposed approach outperforms existing state-of-the-art methods in both quantitative metrics and visual quality. Full article

Electroencephalography (EEG) has attracted significant attention as an effective modality for interaction between the physical and virtual worlds, with EEG-based person identification serving as a key gateway to such applications. Despite substantial progress in EEG-based person identification, several challenges remain: (1) how to design an end-to-end EEG-based identification pipeline; (2) how to perform automatic electrode selection for each user to reduce redundancy and improve discriminative capacity; (3) how to enhance the backbone network’s feature extraction capability by suppressing irrelevant information and better leveraging informative patterns; and (4) how to leverage higher-level information in EEG signals to achieve intent recognition (i.e., EEG-based task/activity recognition under controlled paradigms) on top of person identification. To address these issues, this article proposes, for the first time, a unified deep learning framework that integrates automatic electrode selection, person identification, and intent recognition. We introduce a novel backbone network, AES-MBE, which integrates automatic electrode selection (AES) and intent recognition. The network combines a channel-attention mechanism with a multi-scale bidirectional encoder (MBE), enabling adaptive capture of fine-grained local features while modeling global temporal dependencies in both forward and backward directions. We validate our approach using the PhysioNet EEG Motor Movement/Imagery Dataset (EEGMMIDB), which contains EEG recordings from 109 subjects performing 4 tasks. Compared with state-of-the-art methods, our framework achieves superior performance. Specifically, our method attains a person identification accuracy of 98.82% using only 4 electrodes and an average intent recognition accuracy of 91.58%. In addition, our approach demonstrates strong stability and robustness as the number of users varies, offering insights for future research and practical applications. Full article

In the modern era, demand forecasting enhances the decision-making tasks of industries for controlling production planning and reducing inventory costs. However, the dynamic nature of the fashion and apparel retail industry necessitates precise demand forecasting to optimize supply chain operations and meet customer expectations. Consequently, this research proposes the Formicary Zebra Optimization-Based Distributed Attention-Guided Convolutional Recurrent Neural Network (FZ-DACR) model for improving the demand forecasting. In the proposed approach, the combination of the Formicary Zebra Optimization and Distributed Attention mechanism enabled deep learning architectures to assist in capturing the complex patterns of the retail sales data. Specifically, the neural networks, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs), facilitate extracting the local features and temporal dependencies to analyze the volatile demand patterns. Furthermore, the proposed model integrates visual and textual data to enhance forecasting accuracy. By leveraging the adaptive optimization capabilities of the Formicary Zebra Algorithm, the proposed model effectively extracts features from product images and historical sales data while addressing the complexities of volatile demand patterns. Based on extensive experimental analysis of the proposed model using diverse datasets, the FZ-DACR model achieves superior performance, with minimum error values including MAE of 1.34, MSE of 4.7, RMS of 2.17, and R² of 93.3% using the DRESS dataset. Moreover, the findings highlight the ability of the proposed model in managing the fluctuating trends and supporting inventory and pricing strategies effectively. This innovative approach has significant implications for retailers, enabling more agile supply chains and improved decision making in a highly competitive market. Full article