Search Results (827)

The state monitoring of power load equipment plays a crucial role in ensuring its normal operation. However, in densely deployed environments, the target equipment often exhibits low clarity, making real-time warnings challenging. In this study, a load equipment segmentation and assessment method based on multi-source tensor feature fusion (LSA-MT) is proposed. First, a lightweight residual block based on the attention mechanism is introduced into the backbone network to emphasize key features of load devices and enhance target segmentation efficiency. Second, a 3D edge detail feature perception module is designed to facilitate multi-scale feature fusion while preserving boundary detail features of different devices, thereby improving local recognition accuracy. Finally, tensor decomposition and reorganization are employed to guide visual feature reconstruction in conjunction with equipment monitoring images, while tensor mapping of equipment monitoring data is utilized for automated fault classification. The experimental results demonstrate that LSE-MT produces visually clearer segmentations compared to models such as the classic UNet++ and the more recent EGE-UNet when segmenting multiple load devices, achieving Dice and mIoU scores of 92.48 and 92.90, respectively. Regarding classification across the four datasets, the average accuracy can reach 92.92%. These findings fully demonstrate the effectiveness of the LSA-MT method in load equipment fault alarms and grid operation and maintenance. Full article

Owing to the demand for smoke and flame detection in natural scenes, this paper proposes a lightweight deep learning model, SF-YOLO (Smoke and Fire-YOLO), for video smoke and flame detection in such environments. Firstly, YOLOv11 is employed as the backbone network, combined with the C3k2 module based on a two-path residual attention mechanism, and a target detection head frame with an embedded attention mechanism. This combination enhances the response of the unobscured regions to compensate for the feature loss in occluded regions, thereby addressing the occlusion problem in dynamic backgrounds. Then, a two-channel loss function (W-SIoU) based on dynamic tuning and intelligent focusing is designed to enhance loss computation in the boundary regions, thus improving the YOLOv11 model’s ability to recognize targets with ambiguous boundaries. Finally, the algorithms proposed in this paper are experimentally validated using the self-generated dataset S-Firedata and the public smoke and flame virtual dataset M⁴SFWD. These datasets are derived from internet smoke and flame video frame extraction images and open-source smoke and flame dataset images, respectively. The experimental results demonstrate, compared with deep learning models such as YOLOv8, Gold-YOLO, and Faster-RCNN, the SF-YOLO model proposed in this paper is more lightweight and exhibits higher detection accuracy and robustness. The metrics mAP50 and mAP50-95 are improved by 2.5% and 2.4%, respectively, in the self-made dataset S-Firedata, and by 0.7% and 1.4%, respectively, in the publicly available dataset M⁴SFWD. The research presented in this paper provides practical methods for the automatic detection of smoke and flame in natural scenes, which can further enhance the effectiveness of fire monitoring systems. Full article

This study proposes an enhanced network intrusion detection model, 1D-TCN-ResNet-BiGRU-Multi-Head Attention (TRBMA), aimed at addressing the issues of incomplete learning of temporal features and low accuracy in the classification of malicious traffic found in existing models. The TRBMA model utilizes Temporal Convolutional Networks (TCNs) to improve the ResNet18 architecture and incorporates Bidirectional Gated Recurrent Units (BiGRUs) and Multi-Head Self-Attention mechanisms to enhance the comprehensive learning of temporal features. Additionally, the ResNet network is adapted into a one-dimensional version that is more suitable for processing time-series data, while the AdamW optimizer is employed to improve the convergence speed and generalization ability during model training. Experimental results on the CIC-IDS-2017 dataset indicate that the TRBMA model achieves an accuracy of 98.66% in predicting malicious traffic types, with improvements in precision, recall, and F1-score compared to the baseline model. Furthermore, to address the challenge of low identification rates for malicious traffic types with small sample sizes in unbalanced datasets, this paper introduces TRBMA (BS-OSS), a variant of the TRBMA model that integrates Borderline SMOTE-OSS hybrid sampling. Experimental results demonstrate that this model effectively identifies malicious traffic types with small sample sizes, achieving an overall prediction accuracy of 99.88%, thereby significantly enhancing the performance of the network intrusion detection model. Full article

In the realm of apple cultivation, the efficient and real-time monitoring of Gray Leaf Spot is the foundation of the effective management of pest control, reducing pesticide dependence and easing the burden on the environment. Additionally, it promotes the harmonious development of the agricultural economy and ecological balance. However, due to the dense foliage and diverse lesion characteristics, monitoring the disease faces unprecedented technical challenges. This paper proposes a detection model for Gray Leaf Spot on apple, which is based on an enhanced YOLOv8 network. The details are as follows: (1) we introduce Dynamic Residual Blocks (DRBs) to boost the model’s ability to extract lesion features, thereby improving detection accuracy; (2) add a Self-Balancing Attention Mechanism (SBAY) to optimize the feature fusion and improve the ability to deal with complex backgrounds; and (3) incorporate an ultra-small detection head and simplify the computational model to reduce the complexity of the YOLOv8 network while maintaining the high precision of detection. The experimental results show that the enhanced model outperforms the original YOLOv8 network in detecting Gray Leaf Spot. Notably, when the Intersection over Union (IoU) is 0.5, an improvement of 7.92% in average precision is observed. Therefore, this advanced detection technology holds pivotal significance in advancing the sustainable development of the apple industry and environment-friendly agriculture. Full article

Researchers encounter substantial challenges in medical image classification, mainly due to limited image resolution and low signal-to-noise ratios. This situation makes it difficult for deep learning algorithms to identify abnormal regions based solely on image content accurately. This paper proposes ResGDANet (Residual Group Dual-Channel Attention Network), an enhanced architecture that builds upon ResGANet by incorporating two novel modules: a Dual-Channel Attention Fusion (DCAF) module and a Retention-Memory Transformer (RMT) module. The DCAF module utilizes a dual-path architecture that integrates global average pooling and max pooling operations, effectively enhancing local feature representation through the fusion of channel-wise and spatial attention mechanisms. The RMT module enhances rotation-invariant feature extraction by integrating the retention mechanism from Retentive Networks and the global context modeling capabilities of Vision Transformers. Extensive experiments on the COVID19-CT and ISIC2018 datasets demonstrate the superiority of ResGDANet, achieving classification accuracies of 83.74% and 81.73% respectively, outperforming state-of-the-art models including ResGANet, GvT, and SENet. Ablation studies and visualization analyses further validate the efficacy of the proposed attention module, showing notable enhancements in feature representation capability and classification accuracy. By introducing a more robust and precise classification framework, this research contributes importantly to the progress in medical image analysis. Full article

Aliphatic unsegmented polyurethanes (PUs) have garnered relatively limited attention in the literature, despite their valuable properties such as UV resistance and biocompatibility, making them suitable for biomedical applications. This study focuses on synthesizing poly(ester-urethanes) (PEUs) using 1,6-hexamethylene diisocyanate and the macrodiol α,ω-hydroxy telechelic poly(ε-caprolactone) (HOPCLOH). To optimize the synthesis, a statistical experimental design approach was employed, a methodology not commonly utilized in polymer science. The influence of reaction temperature, time, reagent concentrations, and solvent type on the resulting PEUs was investigated. Characterization techniques included FT-IR, ¹H NMR, differential scanning calorimetry (DSC), gel permeation chromatography (GPC), optical microscopy, and mechanical testing. The results demonstrated that all factors significantly impacted the number-average molecular weight (M_n) as determined by GPC. Furthermore, the statistical design revealed crucial interaction effects between factors, such as a dependence between reaction time and temperature. For example, a fixed reaction time of 1 h, with the temperature varying from 50 °C to 61 °C, did not significantly alter M_n. Better reaction conditions yielded high M_n (average: 162,000 g/mol), desirable mechanical properties (elongation at break > 1000%), low levels of unreacted HOPCLOH in the PEU films (OH/ESTER response = 0.0008), and reduced crystallinity (ΔH_m = 11 J/g) in the soft segment, as observed by DSC and optical microscopy. In contrast, suboptimal conditions resulted in low M_n, brittle materials with unmeasurable mechanical properties, high crystallinity, and significant amounts of residual HOPCLOH. The best experimental conditions were 61 °C, 0.176 molal, 8 h, and chloroform as the solvent (ε = 4.8). Full article

This study presents PolypeptideDesigner (PPD), a novel conditional diffusion-based model for de novo polypeptide sequence design and generation based on per-residue secondary structure conditions. By integrating a lightweight LSTM-attention neural network as the denoiser within a diffusion framework, PPD offers an innovative and efficient approach to polypeptide generation. Evaluations demonstrate that the PPD model can generate diverse and novel polypeptide sequences across various testing conditions, achieving high pLDDT scores when folded by ESMFold. In comparison to the ProteinDiffusionGenerator B (PDG-B) model, a relevant benchmark in the field, PPD exhibits the ability to produce longer and more diverse polypeptide sequences. This improvement is attributed to PPD’s optimized architecture and expanded training dataset, which enhance its understanding of protein structural pattern. The PPD model shows significant potential for optimizing functional polypeptides with known structures, paving the way for advancements in biomaterial design. Future work will focus on further refining the model and exploring its broader applications in polypeptide engineering. Full article

The use of yellow sticky traps is a green pest control method that utilizes the pests’ attraction to the color yellow. The use of yellow sticky traps not only controls pest populations but also enables monitoring, offering a more economical and environmentally friendly alternative to pesticides. However, the small size and dense distribution of pests on yellow sticky traps lead to lower detection accuracy when using lightweight models. On the other hand, large models suffer from longer training times and deployment difficulties, posing challenges for pest detection in the field using edge computing platforms. To address these issues, this paper proposes a lightweight detection method, YOLO-YSTs, based on an improved YOLOv10n model. The method aims to balance pest detection accuracy and model size and has been validated on edge computing platforms. This model incorporates SPD-Conv convolutional modules, the iRMB inverted residual block attention mechanism, and the Inner-SIoU loss function to improve the YOLOv10n network architecture, ultimately addressing the issues of missed and false detections for small and overlapping targets while balancing model speed and accuracy. Experimental results show that the YOLO-YSTs model achieved precision, recall, mAP₅₀, and mAP_50–95 values of 83.2%, 83.2%, 86.8%, and 41.3%, respectively, on the yellow sticky trap dataset. The detection speed reached 139 FPS, with GFLOPs at only 8.8. Compared with the YOLOv10n model, the mAP₅₀ improved by 1.7%. Compared with other mainstream object detection models, YOLO-YSTs also achieved the best overall performance. Through improvements to the YOLOv10n model, the accuracy of pest detection on yellow sticky traps was effectively enhanced, and the model demonstrated good detection performance when deployed on edge mobile platforms. In conclusion, the proposed YOLO-YSTs model offers more balanced performance in the detection of pest images on yellow sticky traps. It performs well when deployed on edge mobile platforms, making it of significant importance for field pest monitoring and integrated pest management. Full article

Traditional sweet potato disease and pest detection methods have the limitations of low efficiency, poor accuracy and manual dependence, while deep learning-based target detection can achieve an efficient and accurate detection. This paper proposed an efficient sweet potato leaf disease and pest detection method SPLDPvB, as well as a low-complexity version SPLDPvT, to achieve accurate identification of sweet potato leaf spots and pests, such as hawk moth and wheat moth. First, a residual module containing three depthwise separable convolutional layers and a skip connection was proposed to effectively retain key feature information. Then, an efficient feature extraction module integrating the residual module and the attention mechanism was designed to significantly improve the feature extraction capability. Finally, in the model architecture, only the structure of the backbone network and the decoupling head combination was retained, and the traditional backbone network was replaced by an efficient feature extraction module, which greatly reduced the model complexity. The experimental results showed that the mAP0.5 and mAP0.5:0.95 of the proposed SPLDPvB model were 88.7% and 74.6%, respectively, and the number of parameters and the amount of calculation were 1.1 M and 7.7 G, respectively. Compared with YOLOv11S, mAP0.5 and mAP0.5:0.95 increased by 2.3% and 2.8%, respectively, and the number of parameters and the amount of calculation were reduced by 88.2% and 63.8%, respectively. The proposed model achieves higher detection accuracy with significantly reduced complexity, demonstrating excellent performance in detecting sweet potato leaf pests and diseases. This method realizes the automatic detection of sweet potato leaf pests and diseases and provides technical guidance for the accurate identification and spraying of pests and diseases. Full article

Ground-based telescopes often encounter challenges such as stray light and vignetting when capturing space objects, leading to non-uniform image backgrounds. This not only weakens the signal-to-noise ratio for target tracking but also reduces the accuracy of recognition systems. To address this challenge, We have proposed a novel network architecture called CycleMRSNet, which is based on the CycleGAN framework and incorporates a multi-scale attention mechanism to enhance image processing capabilities. Specifically, we have introduced a multi-scale feature extraction module (MSFEM) at the front end of the generator and embedded an efficient multi-scale attention residual block (EMA-residual block) within the Resnet backbone network. This design improves the efficiency of feature extraction and increases the focus on multi-scale information in high-dimensional feature maps, enabling the network to more comprehensively understand and concentrate on key areas within images, thereby capably correcting non-uniform backgrounds. To evaluate the performance of CycleMRSNet, we trained the model using a small-scale dataset and conducted corrections on simulated and real images within the test set. Experimental results showed that our model achieved scores of PSNR 32.7923, SSIM 0.9814, and FID 1.9212 in the test set, outperforming other methods. These metrics suggest that our approach significantly improves the correction of non-uniform backgrounds and enhances the robustness of the system. Full article

Surface defect detection is essential for ensuring the quality and safety of steel products. While Transformer-based methods have achieved state-of-the-art performance, they face several limitations, including high computational costs due to the quadratic complexity of the attention mechanism, inadequate detection accuracy for small-scale defects due to substantial downsampling, inconsistencies between classification scores and localization confidence, and feature resolution loss caused by simple upsampling and downsampling strategies. To address these challenges, we propose the HCT-Det model, which incorporates a window-based self-attention residual (WSA-R) block structure. This structure combines window-based self-attention (WSA) blocks to reduce computational overhead and parallel residual convolutional (Res) blocks to enhance local feature continuity. The model’s backbone generates three cross-scale features as encoder inputs, which undergo Intra-Scale Feature Interaction (ISFI) and Cross-Scale Feature Interaction (CSFI) to improve detection accuracy for targets of various sizes. A Soft IoU-Aware mechanism ensures alignment between classification scores and intersection-over-union (IoU) metrics during training. Additionally, Hybrid Downsampling (HDownsample) and Hybrid Upsampling (HUpsample) modules minimize feature degradation. Our experiments demonstrate that HCT-Det achieved a mean average precision ([email protected]) of 0.795 on the NEU-DET dataset and 0.733 on the GC10-DET dataset, outperforming other state-of-the-art approaches. These results highlight the model’s effectiveness in improving computational efficiency and detection accuracy for steel surface defect detection. Full article

Liver cancer is a major global health challenge, significantly contributing to mortality rates. The accurate segmentation of liver and tumors from clinical CT images plays a crucial role in selecting therapeutic strategies for liver disease and treatment monitoring but remains challenging due to liver shape variability, proximity to other organs, low contrast between tumors and healthy tissues, and unclear lesion boundaries. To address these challenges, we propose the Deep Residual Dual-Attention Network (DRDA-Net), a novel model for end-to-end liver and tumor segmentation. DRDA-Net integrates a Residual UNet architecture with dual-attention mechanisms, multi-scale tile and patch extraction, and an Ensemble method. The dual-attention mechanisms enhance focus on key regions, addressing variations in size, shape, and intensity, while the multi-scale approach captures fine details and broader contexts. Additionally, we introduce a unique pre-processing pipeline employing a two-channel denoising technique using convolutional neural networks (CNNs) and stationary wavelet transforms (SWTs) to reduce noise while preserving structural details. Evaluated on the 3DIRCADb dataset, DRDA-Net achieved Dice scores of 97.03% and 75.4% for liver and tumor segmentation, respectively, outperforming state-of-the-art methods. These results demonstrate the model’s effectiveness in overcoming segmentation challenges and highlight its potential to improve liver cancer diagnostics and treatment planning. Full article

Audio tagging, as a fundamental task in acoustic signal processing, has demonstrated significant advances and broad applications in recent years. Spiking Neural Networks (SNNs), inspired by biological neural systems, exploit event-driven computing paradigms and temporal information processing, enabling superior energy efficiency. Despite the increasing adoption of SNNs, the potential of event-driven encoding mechanisms for audio tagging remains largely unexplored. This work presents a pioneering investigation into event-driven encoding strategies for SNN-based audio tagging. We propose the SATRN (Spiking Audio Tagging Robust Network), a novel architecture that integrates temporal–spatial attention mechanisms with membrane potential residual connections. The network employs a dual-stream structure combining global feature fusion and local feature extraction through inverted bottleneck blocks, specifically designed for efficient audio processing. Furthermore, we introduce an event-based encoding approach that enhances the resilience of Spiking Neural Networks to disturbances while maintaining performance. Our experimental results on the Urbansound8k and FSD50K datasets demonstrate that the SATRN achieves comparable performance to traditional Convolutional Neural Networks (CNNs) while requiring significantly less computation time and showing superior robustness against noise perturbations, making it particularly suitable for edge computing scenarios and real-time audio processing applications. Full article

Many important plants’ agronomic traits, such as crop yield, stress tolerance, and other traits, are controlled by multiple genes and exhibit complex inheritance patterns. Traditional breeding methods often encounter difficulties in dealing with these traits due to their complexity. However, genomic selection (GS), which utilizes high-density molecular markers across the entire genome to facilitate selection in breeding programs, excels in capturing the genetic variation associated with these traits. This enables more accurate and efficient selection in breeding. The traditional crop genome selection model, based on statistical methods or machine learning models, often treats samples as independent entities while neglecting the abundance latent relational information among them. Consequently, this limitation hampers their predictive performance. In this study, we proposed a novel crop genome selection model based on hypergraph attention networks for genomic prediction (HGATGS). This model incorporates dynamic hyperedges that are designed based on sample similarity to validate the efficacy of high-order relationships between samples for phenotypic prediction. By introducing an attention mechanism, it assigns weights to different hyperedges and nodes, thereby enhancing the ability to capture kinship relationships among samples. Additionally, residual connections are incorporated between hypergraph convolutional layers to further improve model stability and performance. The model was validated on datasets for multiple crops, including wheat, corn, and rice. The results showed that HGATGS significantly outperformed traditional statistical methods and machine learning models on the Wheat 599, Rice 299, and G2F 2017 datasets. On Wheat 599, HGATGS achieved a correlation coefficient of 0.54, a 14.9% improvement over methods like R-BLUP and BayesA (0.47). On Rice 299, HGATGS reached 0.45, a 66.7% increase compared to other models like R-BLUP and SVR (0.27). On G2F 2017, HGATGS attained 0.88, slightly surpassing other models like R-BLUP and BayesA (0.87). We conducted ablation experiments to compare the model’s performance across three datasets, and found that the model integrating hypergraph attention and residual connections performed optimally. Subsequent comparisons of the model’s prediction performance with dynamically selected different k values revealed optimal performance when K = (3,4). The model’s prediction performance was also compared across different single nucleotide polymorphisms (SNPs) and sample sizes in various datasets, with HGATGS consistently outperforming the comparison models. Finally, visualizations of the constructed hypergraph structures showed that certain nodes have high connection densities with hyperedges. These nodes often represent varieties or genotypes with significant impacts on traits. During feature aggregation, these high-connectivity nodes contribute significantly to the prediction results and demonstrate better prediction performance across multiple traits in multiple crops. This demonstrates that the method of constructing hypergraphs through correlation relationships for prediction is highly effective. Full article

In the field of wind turbine spindle bearing fault diagnosis, real-time monitoring of its operation is challenging. The state monitoring visualization is limited, fault data and sample labels are scarce, and fault data distribution varies under different operational conditions, leading to low diagnosis accuracy and slow diagnosis speed. To address these challenges, a wind turbine spindle bearing fault diagnosis method based on Unity 3D and transfer learning is proposed. Based on the characteristics of the wind turbine spindle bearing structure and operation, a digital twin model is established. The twin data transmit the necessary information to each module in various file formats. Additionally, the signal processing method (RB), combined with a random convolution layer and blind deconvolution, is employed to enhance the diversity of fault features. The processed signal is then fed into an improved residual network model with an efficient channel attention mechanism. Finally, the model incorporates related alignment and joint maximum mean difference for fault diagnosis. This model not only improves the extraction of key features but also adapts to edge and condition distributions through domain adaptation, enabling cross-domain identification. The digital twin system is implemented in Unity 3D, incorporating functions such as user login, wind turbine spindle bearing state monitoring, fault diagnosis, and fault warning, demonstrating practical applicability. Experimental results validate the effectiveness and superiority of the proposed method in fault diagnosis across various transfer learning tasks. Full article