An improved analytical design to investigate the static stiffness of a convoluted air spring is developed and presented in this article.An air spring provides improved ride comfort by achieving variable volume at vari...An improved analytical design to investigate the static stiffness of a convoluted air spring is developed and presented in this article.An air spring provides improved ride comfort by achieving variable volume at various strokes of the suspension.An analytical relation is derived to calculate the volume and the rate of change in the volume of the convoluted bellow with respect to various suspension heights.This expression is used in the equation to calculate the variable stiffness of the bellow.The obtained analytical characteristics are validated with a detailed experiment to test the static vertical stiffness of the air spring.The convoluted air bellow is tested in an Avery spring-testing apparatus for various loads.The bellow is modeled in the ABAQUS environment to perform finite element analysis(FEA)to understand and visualize the deflection of the bellow at various elevated internal pressures and external loads.The proposed air spring model is a fiber-reinforced rubber bellow enclosed between two metal plates.The Mooney-Rivlin material model was used to model the hyperelastic rubber material for FEA.From the results,it is observed that the experimental and analytical results match with a minor error of 7.54%.The derived relations and validations would provide design guidance at the developmental stage of air bellows.These expressions would also play a major role in designing an effective active air suspension system by accurately calculating the required stiffness at various loads.展开更多
The post-hole convolute(PHC),which is used to transport and combine the pulse power flux,is a key component in huge pulsed power generators.Current loss at the PHC is a challenging problem for researchers.To explore a...The post-hole convolute(PHC),which is used to transport and combine the pulse power flux,is a key component in huge pulsed power generators.Current loss at the PHC is a challenging problem for researchers.To explore a method of reducing the current loss,a single-hole PHC was designed for experiments on the current loss on the Qiang Guang I generator.The experimental results showed that the current loss at the single-hole PHC is related to the distance/between the vicinity of the cathode hole and the surface of the downstream side of the post.Meanwhile,a single-hole PHC with a blob cathode hole transmitted current more effectively than the PHC with a circle cathode hole.The relative current loss at the single-hole PHC with the cathode coaled w ith gold foil was about 30%-50% of that with the cathode coated with nickel and titanium foil.The gap closing speed was also obtained from the current waveforms in the experiments.The speed was 5.74-14.52 cmμs 1 which was different from the classical plasma expansion velocity of 3 cmμs 1.展开更多
Reliable traffic flow prediction is crucial for mitigating urban congestion.This paper proposes Attentionbased spatiotemporal Interactive Dynamic Graph Convolutional Network(AIDGCN),a novel architecture integrating In...Reliable traffic flow prediction is crucial for mitigating urban congestion.This paper proposes Attentionbased spatiotemporal Interactive Dynamic Graph Convolutional Network(AIDGCN),a novel architecture integrating Interactive Dynamic Graph Convolution Network(IDGCN)with Temporal Multi-Head Trend-Aware Attention.Its core innovation lies in IDGCN,which uniquely splits sequences into symmetric intervals for interactive feature sharing via dynamic graphs,and a novel attention mechanism incorporating convolutional operations to capture essential local traffic trends—addressing a critical gap in standard attention for continuous data.For 15-and 60-min forecasting on METR-LA,AIDGCN achieves MAEs of 0.75%and 0.39%,and RMSEs of 1.32%and 0.14%,respectively.In the 60-min long-term forecasting of the PEMS-BAY dataset,the AIDGCN out-performs the MRA-BGCN method by 6.28%,4.93%,and 7.17%in terms of MAE,RMSE,and MAPE,respectively.Experimental results demonstrate the superiority of our pro-posed model over state-of-the-art methods.展开更多
Advances in optical coherence tomography(OCT)technology allow a clear view of the vitreoretinal interface(VRI).The abnormality of the VRI is one of the common symptoms of high myopia,mainly including posterior vitreou...Advances in optical coherence tomography(OCT)technology allow a clear view of the vitreoretinal interface(VRI).The abnormality of the VRI is one of the common symptoms of high myopia,mainly including posterior vitreous detachment(PVD)and epiretinal membrane(ERM).They can cause severe damage to the structure and function of the retina,leading to permanent vision loss.Therefore,fully automated detection of abnormalities at the VRI is crucial for the management of high myopia.This paper presents a DS-YOLOv7 network aimed at accurately identifying abnormalities,including partial PVD,complete PVD,and ERM from retinal OCT images.Built upon the YOLOv7 network,the proposed model integrates the advanced dynamic snake convolution(DSConv)module to capture the curvilinear characteristics of lesions,and the mixture of attention and convolution(ACMix)module to improve the precision and robustness of feature extraction through effective fusion of self-attention mechanisms and convolution.Moreover,the introduction of the efficient complete intersection-over-union(ECIoU)loss function further enhances the coordinate regression capability of the model.Threefold cross-validation on a dataset with 1973 OCT B-scans from 46 patients shows that the DS-YOLOv7 achieved superior performance in vitreoretinal interface abnormality detection,with mAP@0.5 of 0.714,mAP@0.75 of 0.438,and mAP@0.5:0.95 of 0.424.The proposed model can provide an accurate and efficient diagnostic tool for patients with high myopia.展开更多
Malware poses a significant threat to the Internet of Things(IoT).It enables unauthorized access to devices in the IoT environment.The lack of unique architectural standards causes challenges in developing robust malw...Malware poses a significant threat to the Internet of Things(IoT).It enables unauthorized access to devices in the IoT environment.The lack of unique architectural standards causes challenges in developing robust malware detection(MD)models.The existing models demand substantial computational resources.This study intends to build a lightweight MD model to detect anomalies in IoT networks.The authors develop a transformation technique,converting the malware binaries into images.MobileNet V2 is fine-tuned using improved grey wolf optimization(IGWO)to extract crucial features of malicious and benign samples.The ResNeXt model is combined with the Linformer’s attention mechanism to identify Malware features.A fully connected layer is integrated with gradientweighted class activation mapping(Grad-CAM)in order to facilitate an interpretable classification model.The proposed model is evaluated using the IoT malware and the IoT-23 datasets.The model performs well on the two datasets with an accuracy of 98.94%,precision of 98.46%,recall of 98.11%,and F1-score of 98.28%on the IoT malware dataset,and an accuracy of 98.23%,precision of 96.80%,recall of 96.64%,and F1-score of 96.71%on the IoT-23 dataset,respectively.The findings indicate that the model has a high standard of classification.The lightweight architecture enables efficient deployment with an inference time of 1.42 s.Inference time has no direct impact on accuracy,precision,recall,or F1-score.However,the inference speed would warrant timely detection in latency-sensitive IoT applications.By achieving a remarkable result,the proposed study offers a comprehensive solution:a scalable,interpretable,and computationally efficient MD model for the evolving IoT landscape.展开更多
In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and rec...In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.展开更多
Red chilli powder(RCP)is a versatile spice accepted globally in diverse culinary products due to its distinct pungent characteristics and red colour.The higher market demand makes the spice vulnerable to unethical mix...Red chilli powder(RCP)is a versatile spice accepted globally in diverse culinary products due to its distinct pungent characteristics and red colour.The higher market demand makes the spice vulnerable to unethical mixing,so its quality assessment is crucial.The non-destructive application of computer vision for measuring food adulteration has always attracted researchers and industry due to its robustness and feasibility.Following the current era of Food Quality 4.0 and artificial intelligence,this study follows an approach based on 1D-convolutional neural networks(CNN)and 2D-CNN models for detecting RCP adulteration.The performance evaluation metrics are used to analyse the efficiency of these models.The histogram features from the Lab colour space trained on the 1D-CNN model(BS-40 and Epoch 100)show an accuracy of 84.56%.On the other hand,the 2D-CNN model DenseNet-121(AdamW and BS-30)also shows a test accuracy of 84.62%.From the observations of this study,it is concluded that CNN models can be a promising tool for solving the adulteration detection problem in food quality evaluation.Further,internet of things-based systems can be developed to aid the industry and government agencies in monitoring the quality of RCP to harness the unethical practices of food adulteration.展开更多
Stereo matching is a pivotal task in computer vision,enabling precise depth estimation from stereo image pairs,yet it encounters challenges in regions with reflections,repetitive textures,or fine structures.In this pa...Stereo matching is a pivotal task in computer vision,enabling precise depth estimation from stereo image pairs,yet it encounters challenges in regions with reflections,repetitive textures,or fine structures.In this paper,we propose a Semantic-Guided Parallax Attention Stereo Matching Network(SGPASMnet)that can be trained in unsupervised manner,building upon the Parallax Attention Stereo Matching Network(PASMnet).Our approach leverages unsupervised learning to address the scarcity of ground truth disparity in stereo matching datasets,facilitating robust training across diverse scene-specific datasets and enhancing generalization.SGPASMnet incorporates two novel components:a Cross-Scale Feature Interaction(CSFI)block and semantic feature augmentation using a pre-trained semantic segmentation model,SegFormer,seamlessly embedded into the parallax attention mechanism.The CSFI block enables effective fusion ofmulti-scale features,integrating coarse and fine details to enhance disparity estimation accuracy.Semantic features,extracted by SegFormer,enrich the parallax attention mechanism by providing high-level scene context,significantly improving performance in ambiguous regions.Our model unifies these enhancements within a cohesive architecture,comprising semantic feature extraction,an hourglass network,a semantic-guided cascaded parallax attentionmodule,outputmodule,and a disparity refinement network.Evaluations on the KITTI2015 dataset demonstrate that our unsupervised method achieves a lower error rate compared to the original PASMnet,highlighting the effectiveness of our enhancements in handling complex scenes.By harnessing unsupervised learning without ground truth disparity needed,SGPASMnet offers a scalable and robust solution for accurate stereo matching,with superior generalization across varied real-world applications.展开更多
The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graph...The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graphs play a crucial role by constructing structured networks of relationships among entities.However,data sparsity and numerous unexplored implicit relations result in the widespread incompleteness of knowledge graphs.In static knowledge graph completion,most existing methods rely on linear operations or simple interaction mechanisms for triple encoding,making it difficult to fully capture the deep semantic associations between entities and relations.Moreover,many methods focus only on the local information of individual triples,ignoring the rich semantic dependencies embedded in the neighboring nodes of entities within the graph structure,which leads to incomplete embedding representations.To address these challenges,we propose Two-Stage Mixer Embedding(TSMixerE),a static knowledge graph completion method based on entity context.In the unit semantic extraction stage,TSMixerE leveragesmulti-scale circular convolution to capture local features atmultiple granularities,enhancing the flexibility and robustness of feature interactions.A channel attention mechanism amplifies key channel responses to suppress noise and irrelevant information,thereby improving the discriminative power and semantic depth of feature representations.For contextual information fusion,a multi-layer self-attentionmechanism enables deep interactions among contextual cues,effectively integrating local details with global context.Simultaneously,type embeddings clarify the semantic identities and roles of each component,enhancing the model’s sensitivity and fusion capabilities for diverse information sources.Furthermore,TSMixerE constructs contextual unit sequences for entities,fully exploring neighborhood information within the graph structure to model complex semantic dependencies,thus improving the completeness and generalization of embedding representations.展开更多
The ubiquity of mobile devices has driven advancements in mobile object detection.However,challenges in multi-scale object detection in open,complex environments persist due to limited computational resources.Traditio...The ubiquity of mobile devices has driven advancements in mobile object detection.However,challenges in multi-scale object detection in open,complex environments persist due to limited computational resources.Traditional approaches like network compression,quantization,and lightweight design often sacrifice accuracy or feature representation robustness.This article introduces the Fast Multi-scale Channel Shuffling Network(FMCSNet),a novel lightweight detection model optimized for mobile devices.FMCSNet integrates a fully convolutional Multilayer Perceptron(MLP)module,offering global perception without significantly increasing parameters,effectively bridging the gap between CNNs and Vision Transformers.FMCSNet achieves a delicate balance between computation and accuracy mainly by two key modules:the ShiftMLP module,including a shift operation and an MLP module,and a Partial group Convolutional(PGConv)module,reducing computation while enhancing information exchange between channels.With a computational complexity of 1.4G FLOPs and 1.3M parameters,FMCSNet outperforms CNN-based and DWConv-based ShuffleNetv2 by 1%and 4.5%mAP on the Pascal VOC 2007 dataset,respectively.Additionally,FMCSNet achieves a mAP of 30.0(0.5:0.95 IoU threshold)with only 2.5G FLOPs and 2.0M parameters.It achieves 32 FPS on low-performance i5-series CPUs,meeting real-time detection requirements.The versatility of the PGConv module’s adaptability across scenarios further highlights FMCSNet as a promising solution for real-time mobile object detection.展开更多
Traditional anomaly detection methods often assume that data points are independent or exhibit regularly structured relationships,as in Euclidean data such as time series or image grids.However,real-world data frequen...Traditional anomaly detection methods often assume that data points are independent or exhibit regularly structured relationships,as in Euclidean data such as time series or image grids.However,real-world data frequently involve irregular,interconnected structures,requiring a shift toward non-Euclidean approaches.This study introduces a novel anomaly detection framework designed to handle non-Euclidean data by modeling transactions as graph signals.By leveraging graph convolution filters,we extract meaningful connection strengths that capture relational dependencies often overlooked in traditional methods.Utilizing the Graph Convolutional Networks(GCN)framework,we integrate graph-based embeddings with conventional anomaly detection models,enhancing performance through relational insights.Ourmethod is validated on European credit card transaction data,demonstrating its effectiveness in detecting fraudulent transactions,particularly thosewith subtle patterns that evade traditional,amountbased detection techniques.The results highlight the advantages of incorporating temporal and structural dependencies into fraud detection,showcasing the robustness and applicability of our approach in complex,real-world scenarios.展开更多
To address the challenge of real-time detection of unauthorized drone intrusions in complex low-altitude urban environments such as parks and airports,this paper proposes an enhanced MBS-YOLO(Multi-Branch Small Target...To address the challenge of real-time detection of unauthorized drone intrusions in complex low-altitude urban environments such as parks and airports,this paper proposes an enhanced MBS-YOLO(Multi-Branch Small Target Detection YOLO)model for anti-drone object detection,based on the YOLOv8 architecture.To overcome the limitations of existing methods in detecting small objects within complex backgrounds,we designed a C2f-Pu module with excellent feature extraction capability and a more compact parameter set,aiming to reduce the model’s computational complexity.To improve multi-scale feature fusion,we construct a Multi-Branch Feature Pyramid Network(MB-FPN)that employs a cross-level feature fusion strategy to enhance the model’s representation of small objects.Additionally,a shared detail-enhanced detection head is introduced to address the large size variations of Unmanned Aerial Vehicle(UAV)targets,thereby improving detection performance across different scales.Experimental results demonstrate that the proposed model achieves consistent improvements across multiple benchmarks.On the Det-Fly dataset,it improves precision by 3%,recall by 5.6%,and mAP50 by 4.5%compared with the baseline,while reducing parameters by 21.2%.Cross-validation on the VisDrone dataset further validates its robustness,yielding additional gains of 3.2%in precision,6.1%in recall,and 4.8%in mAP50 over the original YOLOv8.These findings confirm the effectiveness of the proposed algorithm in enhancing UAV detection performance under complex scenarios.展开更多
Detecting small forest fire targets in unmanned aerial vehicle(UAV)images is difficult,as flames typically cover only a very limited portion of the visual scene.This study proposes Context-guided Compact Lightweight N...Detecting small forest fire targets in unmanned aerial vehicle(UAV)images is difficult,as flames typically cover only a very limited portion of the visual scene.This study proposes Context-guided Compact Lightweight Network(CCLNet),an end-to-end lightweight model designed to detect small forest fire targets while ensuring efficient inference on devices with constrained computational resources.CCLNet employs a three-stage network architecture.Its key components include three modules.C3F-Convolutional Gated Linear Unit(C3F-CGLU)performs selective local feature extraction while preserving fine-grained high-frequency flame details.Context-Guided Feature Fusion Module(CGFM)replaces plain concatenation with triplet-attention interactions to emphasize subtle flame patterns.Lightweight Shared Convolution with Separated Batch Normalization Detection(LSCSBD)reduces parameters through separated batch normalization while maintaining scale-specific statistics.We build TF-11K,an 11,139-image dataset combining 9139 self-collected UAV images from subtropical forests and 2000 re-annotated frames from the FLAME dataset.On TF-11K,CCLNet attains 85.8%mAP@0.5,45.5%mean Average Precision(mAP)@[0.5:0.95],87.4%precision,and 79.1%recall with 2.21 M parameters and 5.7 Giga Floating-point Operations Per Second(GFLOPs).The ablation study confirms that each module contributes to both accuracy and efficiency.Cross-dataset evaluation on DFS yields 77.5%mAP@0.5 and 42.3%mAP@[0.5:0.95],indicating good generalization to unseen scenes.These results suggest that CCLNet offers a practical balance between accuracy and speed for small-target forest fire monitoring with UAVs.展开更多
Mental-health risk detection seeks early signs of distress from social media posts and clinical transcripts to enable timely intervention before crises.When such risks go undetected,consequences can escalate to self-h...Mental-health risk detection seeks early signs of distress from social media posts and clinical transcripts to enable timely intervention before crises.When such risks go undetected,consequences can escalate to self-harm,long-term disability,reduced productivity,and significant societal and economic burden.Despite recent advances,detecting risk from online text remains challenging due to heterogeneous language,evolving semantics,and the sequential emergence of new datasets.Effective solutions must encode clinically meaningful cues,reason about causal relations,and adapt to new domains without forgetting prior knowledge.To address these challenges,this paper presents a Continual Neuro-Symbolic Graph Learning(CNSGL)framework that unifies symbolic reasoning,causal inference,and continual learning within a single architecture.Each post is represented as a symbolic graph linking clinically relevant tags to textual content,enriched with causal edges derived from directional Point-wise Mutual Information(PMI).A two-layer Graph Convolutional Network(GCN)encodes these graphs,and a Transformer-based attention pooler aggregates node embeddings while providing interpretable tag-level importances.Continual adaptation across datasets is achieved through the Multi-Head Freeze(MH-Freeze)strategy,which freezes a shared encoder and incrementally trains lightweight task-specific heads(small classifiers attached to the shared embedding).Experimental evaluations across six diverse mental-health datasets ranging from Reddit discourse to clinical interviews,demonstrate that MH-Freeze consistently outperforms existing continual-learning baselines in both discriminative accuracy and calibration reliability.Across six datasets,MH-Freeze achieves up to 0.925 accuracy and 0.923 F1-Score,with AUPRC≥0.934 and AUROC≥0.942,consistently surpassing all continual-learning baselines.The results confirm the framework’s ability to preserve prior knowledge,adapt to domain shifts,and maintain causal interpretability,establishing CNSGL as a promising step toward robust,explainable,and lifelong mental-health risk assessment.展开更多
The development of brain-computer interfaces(BCI)based on motor imagery(MI)has greatly improved patients’quality of life with movement disorders.The classification of upper limb MI has been widely studied and applied...The development of brain-computer interfaces(BCI)based on motor imagery(MI)has greatly improved patients’quality of life with movement disorders.The classification of upper limb MI has been widely studied and applied in many fields,including rehabilitation.However,the physiological representations of left and right lower limb movements are too close and activated deep in the cerebral cortex,making it difficult to distinguish their features.Therefore,classifying lower limbs motor imagery is more challenging.In this study,we propose a feature extraction method based on functional connectivity,which utilizes phase-locked values to construct a functional connectivity matrix as the features of the left and right legs,which can effectively avoid the problem of physiological representations of the left and right lower limbs being too close to each other during movement.In addition,considering the topology and the temporal characteristics of the electroencephalogram(EEG),we designed a temporal-spatial convolutional network(TSGCN)to capture the spatiotemporal information for classification.Experimental results show that the accuracy of the proposed method is higher than that of existing methods,achieving an average classification accuracy of 73.58%on the internal dataset.Finally,this study explains the network mechanism of left and right foot MI from the perspective of graph theoretic features and demonstrates the feasibility of decoding lower limb MI.展开更多
文摘An improved analytical design to investigate the static stiffness of a convoluted air spring is developed and presented in this article.An air spring provides improved ride comfort by achieving variable volume at various strokes of the suspension.An analytical relation is derived to calculate the volume and the rate of change in the volume of the convoluted bellow with respect to various suspension heights.This expression is used in the equation to calculate the variable stiffness of the bellow.The obtained analytical characteristics are validated with a detailed experiment to test the static vertical stiffness of the air spring.The convoluted air bellow is tested in an Avery spring-testing apparatus for various loads.The bellow is modeled in the ABAQUS environment to perform finite element analysis(FEA)to understand and visualize the deflection of the bellow at various elevated internal pressures and external loads.The proposed air spring model is a fiber-reinforced rubber bellow enclosed between two metal plates.The Mooney-Rivlin material model was used to model the hyperelastic rubber material for FEA.From the results,it is observed that the experimental and analytical results match with a minor error of 7.54%.The derived relations and validations would provide design guidance at the developmental stage of air bellows.These expressions would also play a major role in designing an effective active air suspension system by accurately calculating the required stiffness at various loads.
文摘The post-hole convolute(PHC),which is used to transport and combine the pulse power flux,is a key component in huge pulsed power generators.Current loss at the PHC is a challenging problem for researchers.To explore a method of reducing the current loss,a single-hole PHC was designed for experiments on the current loss on the Qiang Guang I generator.The experimental results showed that the current loss at the single-hole PHC is related to the distance/between the vicinity of the cathode hole and the surface of the downstream side of the post.Meanwhile,a single-hole PHC with a blob cathode hole transmitted current more effectively than the PHC with a circle cathode hole.The relative current loss at the single-hole PHC with the cathode coaled w ith gold foil was about 30%-50% of that with the cathode coated with nickel and titanium foil.The gap closing speed was also obtained from the current waveforms in the experiments.The speed was 5.74-14.52 cmμs 1 which was different from the classical plasma expansion velocity of 3 cmμs 1.
文摘针对果实分拣中存在识别精度低、耗时长等问题,设计实现了一种基于深度学习的智能水果分拣系统.首先,该系统采用残差网络(Residual network,ResNet)模型,通过引入动态残差门控机制优化梯度传播有效解决了深层网络训练中的梯度消失和爆炸问题,使得网络能够通过跳跃连接学习到更有效的特征表示;其次,对ResNet-18模型进行了轻量化设计,利用交叉熵损失函数(CrossEntropy loss,CELoss)和Adam优化器(Adaptive moment estimation,Adam)来进行模型的训练;最后,对数据集peach-split进行实验分析,结果表明构建的智能分拣系统对提高水果分拣精度研究具有一定的实用价值.
文摘Reliable traffic flow prediction is crucial for mitigating urban congestion.This paper proposes Attentionbased spatiotemporal Interactive Dynamic Graph Convolutional Network(AIDGCN),a novel architecture integrating Interactive Dynamic Graph Convolution Network(IDGCN)with Temporal Multi-Head Trend-Aware Attention.Its core innovation lies in IDGCN,which uniquely splits sequences into symmetric intervals for interactive feature sharing via dynamic graphs,and a novel attention mechanism incorporating convolutional operations to capture essential local traffic trends—addressing a critical gap in standard attention for continuous data.For 15-and 60-min forecasting on METR-LA,AIDGCN achieves MAEs of 0.75%and 0.39%,and RMSEs of 1.32%and 0.14%,respectively.In the 60-min long-term forecasting of the PEMS-BAY dataset,the AIDGCN out-performs the MRA-BGCN method by 6.28%,4.93%,and 7.17%in terms of MAE,RMSE,and MAPE,respectively.Experimental results demonstrate the superiority of our pro-posed model over state-of-the-art methods.
基金supported by the National Natural Science Foundation of China(62271337,62371326,and 62371328)the National Key Research and Development Program of China(2019FYC1710204)+1 种基金the National Clinical Key Specialty Construction Project(10000015Z155080000004)the Natural Science Foundation of Jiangsu Province(BK20231310).
文摘Advances in optical coherence tomography(OCT)technology allow a clear view of the vitreoretinal interface(VRI).The abnormality of the VRI is one of the common symptoms of high myopia,mainly including posterior vitreous detachment(PVD)and epiretinal membrane(ERM).They can cause severe damage to the structure and function of the retina,leading to permanent vision loss.Therefore,fully automated detection of abnormalities at the VRI is crucial for the management of high myopia.This paper presents a DS-YOLOv7 network aimed at accurately identifying abnormalities,including partial PVD,complete PVD,and ERM from retinal OCT images.Built upon the YOLOv7 network,the proposed model integrates the advanced dynamic snake convolution(DSConv)module to capture the curvilinear characteristics of lesions,and the mixture of attention and convolution(ACMix)module to improve the precision and robustness of feature extraction through effective fusion of self-attention mechanisms and convolution.Moreover,the introduction of the efficient complete intersection-over-union(ECIoU)loss function further enhances the coordinate regression capability of the model.Threefold cross-validation on a dataset with 1973 OCT B-scans from 46 patients shows that the DS-YOLOv7 achieved superior performance in vitreoretinal interface abnormality detection,with mAP@0.5 of 0.714,mAP@0.75 of 0.438,and mAP@0.5:0.95 of 0.424.The proposed model can provide an accurate and efficient diagnostic tool for patients with high myopia.
基金supported by the Deanship of Scientific Research,Vice Presidency for Graduate Studies and Scientific Research,King Faisal University,Saudi Arabia[Grant No.KFU253774].
文摘Malware poses a significant threat to the Internet of Things(IoT).It enables unauthorized access to devices in the IoT environment.The lack of unique architectural standards causes challenges in developing robust malware detection(MD)models.The existing models demand substantial computational resources.This study intends to build a lightweight MD model to detect anomalies in IoT networks.The authors develop a transformation technique,converting the malware binaries into images.MobileNet V2 is fine-tuned using improved grey wolf optimization(IGWO)to extract crucial features of malicious and benign samples.The ResNeXt model is combined with the Linformer’s attention mechanism to identify Malware features.A fully connected layer is integrated with gradientweighted class activation mapping(Grad-CAM)in order to facilitate an interpretable classification model.The proposed model is evaluated using the IoT malware and the IoT-23 datasets.The model performs well on the two datasets with an accuracy of 98.94%,precision of 98.46%,recall of 98.11%,and F1-score of 98.28%on the IoT malware dataset,and an accuracy of 98.23%,precision of 96.80%,recall of 96.64%,and F1-score of 96.71%on the IoT-23 dataset,respectively.The findings indicate that the model has a high standard of classification.The lightweight architecture enables efficient deployment with an inference time of 1.42 s.Inference time has no direct impact on accuracy,precision,recall,or F1-score.However,the inference speed would warrant timely detection in latency-sensitive IoT applications.By achieving a remarkable result,the proposed study offers a comprehensive solution:a scalable,interpretable,and computationally efficient MD model for the evolving IoT landscape.
文摘In daily life,keyword spotting plays an important role in human-computer interaction.However,noise often interferes with the extraction of time-frequency information,and achieving both computational efficiency and recognition accuracy on resource-constrained devices such as mobile terminals remains a major challenge.To address this,we propose a novel time-frequency dual-branch parallel residual network,which integrates a Dual-Branch Broadcast Residual module and a Time-Frequency Coordinate Attention module.The time-domain and frequency-domain branches are designed in parallel to independently extract temporal and spectral features,effectively avoiding the potential information loss caused by serial stacking,while enhancing information flow and multi-scale feature fusion.In terms of training strategy,a curriculum learning approach is introduced to progressively improve model robustness fromeasy to difficult tasks.Experimental results demonstrate that the proposed method consistently outperforms existing lightweight models under various signal-to-noise ratio(SNR)conditions,achieving superior far-field recognition performance on the Google Speech Commands V2 dataset.Notably,the model maintains stable performance even in low-SNR environments such as–10 dB,and generalizes well to unseen SNR conditions during training,validating its robustness to novel noise scenarios.Furthermore,the proposed model exhibits significantly fewer parameters,making it highly suitable for deployment on resource-limited devices.Overall,the model achieves a favorable balance between performance and parameter efficiency,demonstrating strong potential for practical applications.
文摘Red chilli powder(RCP)is a versatile spice accepted globally in diverse culinary products due to its distinct pungent characteristics and red colour.The higher market demand makes the spice vulnerable to unethical mixing,so its quality assessment is crucial.The non-destructive application of computer vision for measuring food adulteration has always attracted researchers and industry due to its robustness and feasibility.Following the current era of Food Quality 4.0 and artificial intelligence,this study follows an approach based on 1D-convolutional neural networks(CNN)and 2D-CNN models for detecting RCP adulteration.The performance evaluation metrics are used to analyse the efficiency of these models.The histogram features from the Lab colour space trained on the 1D-CNN model(BS-40 and Epoch 100)show an accuracy of 84.56%.On the other hand,the 2D-CNN model DenseNet-121(AdamW and BS-30)also shows a test accuracy of 84.62%.From the observations of this study,it is concluded that CNN models can be a promising tool for solving the adulteration detection problem in food quality evaluation.Further,internet of things-based systems can be developed to aid the industry and government agencies in monitoring the quality of RCP to harness the unethical practices of food adulteration.
基金supported by the National Natural Science Foundation of China,No.62301497the Science and Technology Research Program of Henan,No.252102211024the Key Research and Development Program of Henan,No.231111212000.
文摘Stereo matching is a pivotal task in computer vision,enabling precise depth estimation from stereo image pairs,yet it encounters challenges in regions with reflections,repetitive textures,or fine structures.In this paper,we propose a Semantic-Guided Parallax Attention Stereo Matching Network(SGPASMnet)that can be trained in unsupervised manner,building upon the Parallax Attention Stereo Matching Network(PASMnet).Our approach leverages unsupervised learning to address the scarcity of ground truth disparity in stereo matching datasets,facilitating robust training across diverse scene-specific datasets and enhancing generalization.SGPASMnet incorporates two novel components:a Cross-Scale Feature Interaction(CSFI)block and semantic feature augmentation using a pre-trained semantic segmentation model,SegFormer,seamlessly embedded into the parallax attention mechanism.The CSFI block enables effective fusion ofmulti-scale features,integrating coarse and fine details to enhance disparity estimation accuracy.Semantic features,extracted by SegFormer,enrich the parallax attention mechanism by providing high-level scene context,significantly improving performance in ambiguous regions.Our model unifies these enhancements within a cohesive architecture,comprising semantic feature extraction,an hourglass network,a semantic-guided cascaded parallax attentionmodule,outputmodule,and a disparity refinement network.Evaluations on the KITTI2015 dataset demonstrate that our unsupervised method achieves a lower error rate compared to the original PASMnet,highlighting the effectiveness of our enhancements in handling complex scenes.By harnessing unsupervised learning without ground truth disparity needed,SGPASMnet offers a scalable and robust solution for accurate stereo matching,with superior generalization across varied real-world applications.
基金supported by the National Natural Science Foundation of China(No.62267005)the Chinese Guangxi Natural Science Foundation(No.2023GXNSFAA026493)+1 种基金Guangxi Collaborative Innovation Center ofMulti-Source Information Integration and Intelligent ProcessingGuangxi Academy of Artificial Intelligence.
文摘The rapid development of information technology and accelerated digitalization have led to an explosive growth of data across various fields.As a key technology for knowledge representation and sharing,knowledge graphs play a crucial role by constructing structured networks of relationships among entities.However,data sparsity and numerous unexplored implicit relations result in the widespread incompleteness of knowledge graphs.In static knowledge graph completion,most existing methods rely on linear operations or simple interaction mechanisms for triple encoding,making it difficult to fully capture the deep semantic associations between entities and relations.Moreover,many methods focus only on the local information of individual triples,ignoring the rich semantic dependencies embedded in the neighboring nodes of entities within the graph structure,which leads to incomplete embedding representations.To address these challenges,we propose Two-Stage Mixer Embedding(TSMixerE),a static knowledge graph completion method based on entity context.In the unit semantic extraction stage,TSMixerE leveragesmulti-scale circular convolution to capture local features atmultiple granularities,enhancing the flexibility and robustness of feature interactions.A channel attention mechanism amplifies key channel responses to suppress noise and irrelevant information,thereby improving the discriminative power and semantic depth of feature representations.For contextual information fusion,a multi-layer self-attentionmechanism enables deep interactions among contextual cues,effectively integrating local details with global context.Simultaneously,type embeddings clarify the semantic identities and roles of each component,enhancing the model’s sensitivity and fusion capabilities for diverse information sources.Furthermore,TSMixerE constructs contextual unit sequences for entities,fully exploring neighborhood information within the graph structure to model complex semantic dependencies,thus improving the completeness and generalization of embedding representations.
基金funded by the National Natural Science Foundation of China under Grant No.62371187the Open Program of Hunan Intelligent Rehabilitation Robot and Auxiliary Equipment Engineering Technology Research Center under Grant No.2024JS101.
文摘The ubiquity of mobile devices has driven advancements in mobile object detection.However,challenges in multi-scale object detection in open,complex environments persist due to limited computational resources.Traditional approaches like network compression,quantization,and lightweight design often sacrifice accuracy or feature representation robustness.This article introduces the Fast Multi-scale Channel Shuffling Network(FMCSNet),a novel lightweight detection model optimized for mobile devices.FMCSNet integrates a fully convolutional Multilayer Perceptron(MLP)module,offering global perception without significantly increasing parameters,effectively bridging the gap between CNNs and Vision Transformers.FMCSNet achieves a delicate balance between computation and accuracy mainly by two key modules:the ShiftMLP module,including a shift operation and an MLP module,and a Partial group Convolutional(PGConv)module,reducing computation while enhancing information exchange between channels.With a computational complexity of 1.4G FLOPs and 1.3M parameters,FMCSNet outperforms CNN-based and DWConv-based ShuffleNetv2 by 1%and 4.5%mAP on the Pascal VOC 2007 dataset,respectively.Additionally,FMCSNet achieves a mAP of 30.0(0.5:0.95 IoU threshold)with only 2.5G FLOPs and 2.0M parameters.It achieves 32 FPS on low-performance i5-series CPUs,meeting real-time detection requirements.The versatility of the PGConv module’s adaptability across scenarios further highlights FMCSNet as a promising solution for real-time mobile object detection.
基金supported by the National Research Foundation of Korea(NRF)funded by the Korea government(RS-2023-00249743)Additionally,this research was supported by the Global-Learning&Academic Research Institution for Master’s,PhD Students,and Postdocs(LAMP)Program of the National Research Foundation of Korea(NRF)grant funded by the Ministry of Education(RS-2024-00443714)This research was also supported by the“Research Base Construction Fund Support Program”funded by Jeonbuk National University in 2025.
文摘Traditional anomaly detection methods often assume that data points are independent or exhibit regularly structured relationships,as in Euclidean data such as time series or image grids.However,real-world data frequently involve irregular,interconnected structures,requiring a shift toward non-Euclidean approaches.This study introduces a novel anomaly detection framework designed to handle non-Euclidean data by modeling transactions as graph signals.By leveraging graph convolution filters,we extract meaningful connection strengths that capture relational dependencies often overlooked in traditional methods.Utilizing the Graph Convolutional Networks(GCN)framework,we integrate graph-based embeddings with conventional anomaly detection models,enhancing performance through relational insights.Ourmethod is validated on European credit card transaction data,demonstrating its effectiveness in detecting fraudulent transactions,particularly thosewith subtle patterns that evade traditional,amountbased detection techniques.The results highlight the advantages of incorporating temporal and structural dependencies into fraud detection,showcasing the robustness and applicability of our approach in complex,real-world scenarios.
基金supported by the Key R&D Programof Xianyang City,Shaanxi Province(L2024-ZDYF-ZDYF-GY-0043).
文摘To address the challenge of real-time detection of unauthorized drone intrusions in complex low-altitude urban environments such as parks and airports,this paper proposes an enhanced MBS-YOLO(Multi-Branch Small Target Detection YOLO)model for anti-drone object detection,based on the YOLOv8 architecture.To overcome the limitations of existing methods in detecting small objects within complex backgrounds,we designed a C2f-Pu module with excellent feature extraction capability and a more compact parameter set,aiming to reduce the model’s computational complexity.To improve multi-scale feature fusion,we construct a Multi-Branch Feature Pyramid Network(MB-FPN)that employs a cross-level feature fusion strategy to enhance the model’s representation of small objects.Additionally,a shared detail-enhanced detection head is introduced to address the large size variations of Unmanned Aerial Vehicle(UAV)targets,thereby improving detection performance across different scales.Experimental results demonstrate that the proposed model achieves consistent improvements across multiple benchmarks.On the Det-Fly dataset,it improves precision by 3%,recall by 5.6%,and mAP50 by 4.5%compared with the baseline,while reducing parameters by 21.2%.Cross-validation on the VisDrone dataset further validates its robustness,yielding additional gains of 3.2%in precision,6.1%in recall,and 4.8%in mAP50 over the original YOLOv8.These findings confirm the effectiveness of the proposed algorithm in enhancing UAV detection performance under complex scenarios.
基金funded by the Natural Science Foundation of Hunan Province(Grant No.2025JJ80352)the National Natural Science Foundation Project of China(Grant No.32271879).
文摘Detecting small forest fire targets in unmanned aerial vehicle(UAV)images is difficult,as flames typically cover only a very limited portion of the visual scene.This study proposes Context-guided Compact Lightweight Network(CCLNet),an end-to-end lightweight model designed to detect small forest fire targets while ensuring efficient inference on devices with constrained computational resources.CCLNet employs a three-stage network architecture.Its key components include three modules.C3F-Convolutional Gated Linear Unit(C3F-CGLU)performs selective local feature extraction while preserving fine-grained high-frequency flame details.Context-Guided Feature Fusion Module(CGFM)replaces plain concatenation with triplet-attention interactions to emphasize subtle flame patterns.Lightweight Shared Convolution with Separated Batch Normalization Detection(LSCSBD)reduces parameters through separated batch normalization while maintaining scale-specific statistics.We build TF-11K,an 11,139-image dataset combining 9139 self-collected UAV images from subtropical forests and 2000 re-annotated frames from the FLAME dataset.On TF-11K,CCLNet attains 85.8%mAP@0.5,45.5%mean Average Precision(mAP)@[0.5:0.95],87.4%precision,and 79.1%recall with 2.21 M parameters and 5.7 Giga Floating-point Operations Per Second(GFLOPs).The ablation study confirms that each module contributes to both accuracy and efficiency.Cross-dataset evaluation on DFS yields 77.5%mAP@0.5 and 42.3%mAP@[0.5:0.95],indicating good generalization to unseen scenes.These results suggest that CCLNet offers a practical balance between accuracy and speed for small-target forest fire monitoring with UAVs.
基金supported by the National Research Foundation of Korea(NRF)grant funded by the Korea government(MSIT)(RS-2025-00518960)in part by the National Research Foundation of Korea(NRF)grant funded by the Korea government(MSIT)(RS-2025-00563192).
文摘Mental-health risk detection seeks early signs of distress from social media posts and clinical transcripts to enable timely intervention before crises.When such risks go undetected,consequences can escalate to self-harm,long-term disability,reduced productivity,and significant societal and economic burden.Despite recent advances,detecting risk from online text remains challenging due to heterogeneous language,evolving semantics,and the sequential emergence of new datasets.Effective solutions must encode clinically meaningful cues,reason about causal relations,and adapt to new domains without forgetting prior knowledge.To address these challenges,this paper presents a Continual Neuro-Symbolic Graph Learning(CNSGL)framework that unifies symbolic reasoning,causal inference,and continual learning within a single architecture.Each post is represented as a symbolic graph linking clinically relevant tags to textual content,enriched with causal edges derived from directional Point-wise Mutual Information(PMI).A two-layer Graph Convolutional Network(GCN)encodes these graphs,and a Transformer-based attention pooler aggregates node embeddings while providing interpretable tag-level importances.Continual adaptation across datasets is achieved through the Multi-Head Freeze(MH-Freeze)strategy,which freezes a shared encoder and incrementally trains lightweight task-specific heads(small classifiers attached to the shared embedding).Experimental evaluations across six diverse mental-health datasets ranging from Reddit discourse to clinical interviews,demonstrate that MH-Freeze consistently outperforms existing continual-learning baselines in both discriminative accuracy and calibration reliability.Across six datasets,MH-Freeze achieves up to 0.925 accuracy and 0.923 F1-Score,with AUPRC≥0.934 and AUROC≥0.942,consistently surpassing all continual-learning baselines.The results confirm the framework’s ability to preserve prior knowledge,adapt to domain shifts,and maintain causal interpretability,establishing CNSGL as a promising step toward robust,explainable,and lifelong mental-health risk assessment.
基金supported in part by the National Natural Science Foundation of China under Grant 62172368the Natural Science Foundation of Zhejiang Province under Grant LR22F020003.
文摘The development of brain-computer interfaces(BCI)based on motor imagery(MI)has greatly improved patients’quality of life with movement disorders.The classification of upper limb MI has been widely studied and applied in many fields,including rehabilitation.However,the physiological representations of left and right lower limb movements are too close and activated deep in the cerebral cortex,making it difficult to distinguish their features.Therefore,classifying lower limbs motor imagery is more challenging.In this study,we propose a feature extraction method based on functional connectivity,which utilizes phase-locked values to construct a functional connectivity matrix as the features of the left and right legs,which can effectively avoid the problem of physiological representations of the left and right lower limbs being too close to each other during movement.In addition,considering the topology and the temporal characteristics of the electroencephalogram(EEG),we designed a temporal-spatial convolutional network(TSGCN)to capture the spatiotemporal information for classification.Experimental results show that the accuracy of the proposed method is higher than that of existing methods,achieving an average classification accuracy of 73.58%on the internal dataset.Finally,this study explains the network mechanism of left and right foot MI from the perspective of graph theoretic features and demonstrates the feasibility of decoding lower limb MI.
