Dynamic sign language recognition holds significant importance, particularly with the application of deep learning to address its complexity. However, existing methods face several challenges. Firstly, recognizing dyn...Dynamic sign language recognition holds significant importance, particularly with the application of deep learning to address its complexity. However, existing methods face several challenges. Firstly, recognizing dynamic sign language requires identifying keyframes that best represent the signs, and missing these keyframes reduces accuracy. Secondly, some methods do not focus enough on hand regions, which are small within the overall frame, leading to information loss. To address these challenges, we propose a novel Video Transformer Attention-based Network (VTAN) for dynamic sign language recognition. Our approach prioritizes informative frames and hand regions effectively. To tackle the first issue, we designed a keyframe extraction module enhanced by a convolutional autoencoder, which focuses on selecting information-rich frames and eliminating redundant ones from the video sequences. For the second issue, we developed a soft attention-based transformer module that emphasizes extracting features from hand regions, ensuring that the network pays more attention to hand information within sequences. This dual-focus approach improves effective dynamic sign language recognition by addressing the key challenges of identifying critical frames and emphasizing hand regions. Experimental results on two public benchmark datasets demonstrate the effectiveness of our network, outperforming most of the typical methods in sign language recognition tasks.展开更多
Network embedding aspires to learn a low-dimensional vector of each node in networks,which can apply to diverse data mining tasks.In real-life,many networks include rich attributes and temporal information.However,mos...Network embedding aspires to learn a low-dimensional vector of each node in networks,which can apply to diverse data mining tasks.In real-life,many networks include rich attributes and temporal information.However,most existing embedding approaches ignore either temporal information or network attributes.A self-attention based architecture using higher-order weights and node attributes for both static and temporal attributed network embedding is presented in this article.A random walk sampling algorithm based on higher-order weights and node attributes to capture network topological features is presented.For static attributed networks,the algorithm incorporates first-order to k-order weights,and node attribute similarities into one weighted graph to preserve topological features of networks.For temporal attribute networks,the algorithm incorporates previous snapshots of networks containing first-order to k-order weights,and nodes attribute similarities into one weighted graph.In addition,the algorithm utilises a damping factor to ensure that the more recent snapshots allocate a greater weight.Attribute features are then incorporated into topological features.Next,the authors adopt the most advanced architecture,Self-Attention Networks,to learn node representations.Experimental results on node classification of static attributed networks and link prediction of temporal attributed networks reveal that our proposed approach is competitive against diverse state-of-the-art baseline approaches.展开更多
SAR images commonly suffer fromspeckle noise,posing a significant challenge in their analysis and interpretation.Existing convolutional neural network(CNN)based despeckling methods have shown great performance in remo...SAR images commonly suffer fromspeckle noise,posing a significant challenge in their analysis and interpretation.Existing convolutional neural network(CNN)based despeckling methods have shown great performance in removing speckle noise.However,these CNN-basedmethods have a fewlimitations.They do not decouple complex background information in amulti-resolutionmanner.Moreover,they have deep network structures thatmay result in many parameters,limiting their applicability tomobile devices.Furthermore,extracting key speckle information in the presence of complex background is also a major problem with SAR.The proposed study addresses these limitations by introducing a lightweight pyramid and attention-based despeckling(PAN-Despeck)network.The primary objective is to enhance image quality and enable improved information interpretation,particularly on mobile devices and scenarios involving complex backgrounds.The PAN-Despeck network leverages domainspecific knowledge and integrates Gaussian Laplacian image pyramid decomposition for multi-resolution image analysis.By utilizing this approach,complex background information can be effectively decoupled,leading to enhanced despeckling performance.Furthermore,the attention mechanism selectively focuses on key speckle features and facilitates complex background removal.The network incorporates recursive and residual blocks to ensure computational efficiency and accelerate training speed,making it lightweight while maintaining high performance.Through comprehensive evaluations,it is demonstrated that PAN-Despeck outperforms existing image restoration methods.With an impressive average peak signal-to-noise ratio(PSNR)of 28.355114 and a remarkable structural similarity index(SSIM)of 0.905467,it demonstrates exceptional performance in effectively reducing speckle noise in SAR images.The source code for the PAN-DeSpeck network is available on GitHub.展开更多
In the current digital era,new technologies are becoming an essential part of our lives.Consequently,the number ofmalicious software ormalware attacks is rapidly growing.There is no doubt,themajority ofmalware attacks...In the current digital era,new technologies are becoming an essential part of our lives.Consequently,the number ofmalicious software ormalware attacks is rapidly growing.There is no doubt,themajority ofmalware attacks can be detected by most antivirus programs.However,such types of antivirus programs are one step behind malicious software.Due to these dilemmas,deep learning become popular in the detection and classification of malicious data.Therefore,researchers have significantly focused on finding solutions for malware attacks by analyzing malicious samples with the help of different techniques and models.In this research,we presented a lightweight attention-based novel deep Convolutional Neural Network(DNN-CNN)model for binary and multi-class malware classification,including benign,trojan horse,ransomware,and spyware.We applied the Principal Component Analysis(PCA)technique for feature extraction for binary classification.We used the Synthetic Minority Oversampling Technique(SMOTE)to handle the imbalanced data during multi-class classification.Our proposed attention-based malware detectionmodel is trained on the benchmarkmalware memory dataset named CIC-MalMem-2022.Theresults indicate that our model obtained high accuracy for binary and multi-class classification,99.5% and 97.9%,respectively.展开更多
In this paper,we demonstrate an application of the Transformer self-attention mechanism in the context of materials science.Our network,the Compositionally Restricted Attention-Based network(CrabNet),explores the area...In this paper,we demonstrate an application of the Transformer self-attention mechanism in the context of materials science.Our network,the Compositionally Restricted Attention-Based network(CrabNet),explores the area of structure-agnostic materials property predictions when only a chemical formula is provided.Our results show that CrabNet’s performance matches or exceeds current best-practice methods on nearly all of 28 total benchmark datasets.We also demonstrate how CrabNet’s architecture lends itself towards model interpretability by showing different visualization approaches that are made possible by its design.We feel confident that CrabNet and its attention-based framework will be of keen interest to future materials informatics researchers.展开更多
Traffic flow prediction is an important part of the intelligent transportation system. Accurate multi-step traffic flow prediction plays an important role in improving the operational efficiency of the traffic network...Traffic flow prediction is an important part of the intelligent transportation system. Accurate multi-step traffic flow prediction plays an important role in improving the operational efficiency of the traffic network. Since traffic flow data has complex spatio-temporal correlation and non-linearity, existing prediction methods are mainly accomplished through a combination of a Graph Convolutional Network (GCN) and a recurrent neural network. The combination strategy has an excellent performance in traffic prediction tasks. However, multi-step prediction error accumulates with the predicted step size. Some scholars use multiple sampling sequences to achieve more accurate prediction results. But it requires high hardware conditions and multiplied training time. Considering the spatiotemporal correlation of traffic flow and influence of external factors, we propose an Attention Based Spatio-Temporal Graph Convolutional Network considering External Factors (ABSTGCN-EF) for multi-step traffic flow prediction. This model models the traffic flow as diffusion on a digraph and extracts the spatial characteristics of traffic flow through GCN. We add meaningful time-slots attention to the encoder-decoder to form an Attention Encoder Network (AEN) to handle temporal correlation. The attention vector is used as a competitive choice to draw the correlation between predicted states and historical states. We considered the impact of three external factors (daytime, weekdays, and traffic accident markers) on the traffic flow prediction tasks. Experiments on two public data sets show that it makes sense to consider external factors. The prediction performance of our ABSTGCN-EF model achieves 7.2%–8.7% higher than the state-of-the-art baselines.展开更多
Electrocardiogram(ECG)signal is one of the noninvasive physiological measurement techniques commonly usedin cardiac diagnosis.However,in real scenarios,the ECGsignal is susceptible to various noise erosion,which affec...Electrocardiogram(ECG)signal is one of the noninvasive physiological measurement techniques commonly usedin cardiac diagnosis.However,in real scenarios,the ECGsignal is susceptible to various noise erosion,which affectsthe subsequent pathological analysis.Therefore,the effective removal of the noise from ECG signals has becomea top priority in cardiac diagnostic research.Aiming at the problem of incomplete signal shape retention andlow signal-to-noise ratio(SNR)after denoising,a novel ECG denoising network,named attention-based residualdense shrinkage network(ARDSN),is proposed in this paper.Firstly,the shallow ECG characteristics are extractedby a shallow feature extraction network(SFEN).Then,the residual dense shrinkage attention block(RDSAB)isused for adaptive noise suppression.Finally,feature fusion representation(FFR)is performed on the hierarchicalfeatures extracted by a series of RDSABs to reconstruct the de-noised ECG signal.Experiments on the MIT-BIHarrhythmia database and MIT-BIH noise stress test database indicate that the proposed scheme can effectively resistthe interference of different sources of noise on the ECG signal.展开更多
A correct and timely fault diagnosis is important for improving the safety and reliability of chemical processes. With the advancement of big data technology, data-driven fault diagnosis methods are being extensively ...A correct and timely fault diagnosis is important for improving the safety and reliability of chemical processes. With the advancement of big data technology, data-driven fault diagnosis methods are being extensively used and still have considerable potential. In recent years, methods based on deep neural networks have made significant breakthroughs, and fault diagnosis methods for industrial processes based on deep learning have attracted considerable research attention. Therefore, we propose a fusion deeplearning algorithm based on a fully convolutional neural network(FCN) to extract features and build models to correctly diagnose all types of faults. We use long short-term memory(LSTM) units to expand our proposed FCN so that our proposed deep learning model can better extract the time-domain features of chemical process data. We also introduce the attention mechanism into the model, aimed at highlighting the importance of features, which is significant for the fault diagnosis of chemical processes with many features. When applied to the benchmark Tennessee Eastman process, our proposed model exhibits impressive performance, demonstrating the effectiveness of the attention-based LSTM FCN in chemical process fault diagnosis.展开更多
It is a common fact that data(features,characteristics or variables)are collected at different sampling frequencies in some fields such as economic and industry.The existing methods usually either ignore the differenc...It is a common fact that data(features,characteristics or variables)are collected at different sampling frequencies in some fields such as economic and industry.The existing methods usually either ignore the difference from the different sampling frequencies or hardly take notice of the inherent temporal characteristics in mixed frequency data.The authors propose an innovative dual attention-based neural network for mixed frequency data(MID-DualAtt),in order to utilize the inherent temporal characteristics and select the input characteristics reasonably without losing information.According to the authors’knowledge,this is the first study to use the attention mechanism to process mixed fre-quency data.The MID-DualAtt model uses the frequency alignment method to trans-form the high--frequency variables into observation vectors at low frequency,and more critical input characteristics are selected for the current prediction index by attention mechanism.The temporal characteristics are explored by the encoder-decoder with attention based on long-short-term memory networks(LSTM).The proposed MID-DualAtt has been tested in practical application,and the experimental results show that it has better prediction ability than the compared models.展开更多
Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learn...Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learning method, named Attention-Based Prototypical Network, is proposed for forest fire smoke detection. Specifically, feature extraction network, which consists of convolutional block attention module, could extract high-level and discriminative features and further decrease the false alarm rate resulting from suspected smoke areas. Moreover, we design a metalearning module to alleviate the overfitting issue caused by limited smoke images, and the meta-learning network enables achieving effective detection via comparing the distance between the class prototype of support images and the features of query images. A series of experiments on forest fire smoke datasets and miniImageNet dataset testify that the proposed method is superior to state-of-the-art few-shot learning approaches.展开更多
The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sa...The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sampling rate,how to model longsequence data and make rational use of the relevant information between channels is also an urgent problem to be solved.In order to solve the above problems,the performance of the end-to-end music separation algorithm is enhanced by improving the network structure.Our main contributions include the following:(1)A more reasonable densely connected U-Net is designed to capture the long-term characteristics of music,such as main melody,tone and so on.(2)On this basis,the multi-head attention and dualpath transformer are introduced in the separation module.Channel attention units are applied recursively on the feature map of each layer of the network,enabling the network to perform long-sequence separation.Experimental results show that after the introduction of the channel attention,the performance of the proposed algorithm has a stable improvement compared with the baseline system.On the MUSDB18 dataset,the average score of the separated audio exceeds that of the current best-performing music separation algorithm based on the time-frequency domain(T-F domain).展开更多
Whole brain functional connectivity(FC)patterns obtained from resting-state functional magnetic resonance imaging(rs-fMRI)have been widely used in the diagnosis of brain disorders such as autism spectrum disorder(ASD)...Whole brain functional connectivity(FC)patterns obtained from resting-state functional magnetic resonance imaging(rs-fMRI)have been widely used in the diagnosis of brain disorders such as autism spectrum disorder(ASD).Recently,an increasing number of studies have focused on employing deep learning techniques to analyze FC patterns for brain disease classification.However,the high dimensionality of the FC features and the interpretation of deep learning results are issues that need to be addressed in the FC-based brain disease classification.In this paper,we proposed a multi-scale attention-based deep neural network(MSA-DNN)model to classify FC patterns for the ASD diagnosis.The model was implemented by adding a flexible multi-scale attention(MSA)module to the auto-encoder based backbone DNN,which can extract multi-scale features of the FC patterns and change the level of attention for different FCs by continuous learning.Our model will reinforce the weights of important FC features while suppress the unimportant FCs to ensure the sparsity of the model weights and enhance the model interpretability.We performed systematic experiments on the large multi-sites ASD dataset with both ten-fold and leaveone-site-out cross-validations.Results showed that our model outperformed classical methods in brain disease classification and revealed robust intersite prediction performance.We also localized important FC features and brain regions associated with ASD classification.Overall,our study further promotes the biomarker detection and computer-aided classification for ASD diagnosis,and the proposed MSA module is flexible and easy to implement in other classification networks.展开更多
As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word em...As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word embedding is utilized to extract information from these templates.However,this method neglects a portion of the content within the logs and confronts the challenge of data imbalance among various log template types after parsing.Currently,specialized research on data imbalance across log template categories remains scarce.A dual-attention-based log anomaly detection model(LogDA),which leveraged data imbalance,was proposed to address these issues in the work.The LogDA model initially utilized a pre-trained model to extract semantic embedding from log templates.Besides,the similarity between embedding was calculated to discern the relationships among the various templates.Then,a Transformer model with a dual-attention mechanism was constructed to capture positional information and global dependencies.Compared to multiple baseline experiments across three public datasets,the proposed approach could improve precision,recall,and F1 scores.展开更多
Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as L...Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as LSTM and ARIMA are better than convolutional neural network in time series prediction,but they are not enough to mine the periodicity of data.In this article,we perform periodic analysis on two types of time series data,select time metrics with high periodic characteristics,and propose a multi-scale prediction model based on the attention mechanism for the periodic trend of the data.A loss calculation method for traffic time series characteristics is proposed as well.Multiple experiments have been conducted on actual data sets.The experiments show that the method proposed in this paper has better performance than commonly used traffic prediction methods(ARIMA,LSTM,etc.)and 3%-5%increase on MAPE.展开更多
Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in hum...Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in human actions makes it difficult to recognize with considerable accuracy.This paper presents a novel deep neural network architecture called Attention RB-Net for HAR using video frames.The input is provided to the model in the form of video frames.The proposed deep architecture is based on the unique structuring of residual blocks with several filter sizes.Features are extracted from each frame via several operations with specific parameters defined in the presented novel Attention-based Residual Bottleneck(Attention-RB)DCNN architecture.A fully connected layer receives an attention-based features matrix,and final classification is performed.Several hyperparameters of the proposed model are initialized using Bayesian Optimization(BO)and later utilized in the trained model for testing.In testing,features are extracted from the self-attention layer and passed to neural network classifiers for the final action classification.Two highly cited datasets,HMDB51 and UCF101,were used to validate the proposed architecture and obtained an average accuracy of 87.70%and 97.30%,respectively.The deep convolutional neural network(DCNN)architecture is compared with state-of-the-art(SOTA)methods,including pre-trained models,inside blocks,and recently published techniques,and performs better.展开更多
Traditional Computational Fluid Dynamics(CFD)simulations are computationally expensive when applied to complex fluid–structure interaction problems and often struggle to capture the essential flow features governing ...Traditional Computational Fluid Dynamics(CFD)simulations are computationally expensive when applied to complex fluid–structure interaction problems and often struggle to capture the essential flow features governing vortex-induced vibrations(VIV)of floating structures.To overcome these limitations,this study develops a hybrid framework that integrates high-fidelity CFD modeling with deep learning techniques to enhance the accuracy and efficiency of VIV response prediction.First,an unstructured finite-volume fluid–structure coupling model is established to generate high-resolution flow field data and extract multi-component time-series feature tensors.These tensors serve as inputs to a Squeeze-and-Excitation Convolutional Neural Network(SE-CNN),which models the nonlinear coupling between flow disturbances and structural responses.The SE-CNN architecture incorporates an attention-based weighting mechanism through an embedded Squeeze-and-Excitation module,dynamically optimizing channel feature importance and improving sensitivity to critical flow characteristics.During training,multidimensional inputs,including pressure,velocity gradient,and displacement sequences,are used to capture the full complexity of fluid–structure interactions.Results demonstrate that the proposed method achieves a maximum amplitude prediction error of only 2.9%and a main frequency deviation below 0.03 Hz,outperforming conventional CNN models by reducing amplitude prediction error from 3.2%to 1.9%.The approach is validated using a representative semi-submersible platform,confirming its robustness across varying damping conditions and flow velocities.展开更多
We present a novel approach for the prediction of crystal material properties that is distinct from the computationally complex and expensive density functional theory(DFT)-based calculations.Instead,we utilize an att...We present a novel approach for the prediction of crystal material properties that is distinct from the computationally complex and expensive density functional theory(DFT)-based calculations.Instead,we utilize an attention-based graph neural network that yields high-accuracy predictions.Our approach employs two attention mechanisms that allow for message passing on the crystal graphs,which in turn enable the model to selectively attend to pertinent atoms and their local environments,thereby improving performance.We conduct comprehensive experiments to validate our approach,which demonstrates that our method surpasses existing methods in terms of predictive accuracy.Our results suggest that deep learning,particularly attention-based networks,holds significant promise for predicting crystal material properties,with implications for material discovery and the refined intelligent systems.展开更多
Recommending personalized travel routes from sparse,implicit feedback poses a significant challenge,as conventional systems often struggle with information overload and fail to capture the complex,sequential nature of...Recommending personalized travel routes from sparse,implicit feedback poses a significant challenge,as conventional systems often struggle with information overload and fail to capture the complex,sequential nature of user preferences.To address this,we propose a Conditional Generative Adversarial Network(CGAN)that generates diverse and highly relevant itineraries.Our approach begins by constructing a conditional vector that encapsulates a user’s profile.This vector uniquely fuses embeddings from a Heterogeneous Information Network(HIN)to model complex user-place-route relationships,a Recurrent Neural Network(RNN)to capture sequential path dynamics,and Neural Collaborative Filtering(NCF)to incorporate collaborative signals from the wider user base.This comprehensive condition,further enhanced with features representing user interaction confidence and uncertainty,steers a CGAN stabilized by spectral normalization to generate high-fidelity latent route representations,effectively mitigating the data sparsity problem.Recommendations are then formulated using an Anchor-and-Expand algorithm,which selects relevant starting Points of Interest(POI)based on user history,then expands routes through latent similarity matching and geographic coherence optimization,culminating in Traveling Salesman Problem(TSP)-based route optimization for practical travel distances.Experiments on a real-world check-in dataset validate our model’s unique generative capability,achieving F1 scores ranging from 0.163 to 0.305,and near-zero pairs−F1 scores between 0.002 and 0.022.These results confirm the model’s success in generating novel travel routes by recommending new locations and sequences rather than replicating users’past itineraries.This work provides a robust solution for personalized travel planning,capable of generating novel and compelling routes for both new and existing users by learning from collective travel intelligence.展开更多
The advent of sixth-generation(6G)networks introduces unprecedented challenges in achieving seamless connectivity,ultra-low latency,and efficient resource management in highly dynamic environments.Although fifth-gener...The advent of sixth-generation(6G)networks introduces unprecedented challenges in achieving seamless connectivity,ultra-low latency,and efficient resource management in highly dynamic environments.Although fifth-generation(5G)networks transformed mobile broadband and machine-type communications at massive scales,their properties of scaling,interference management,and latency remain a limitation in dense high mobility settings.To overcome these limitations,artificial intelligence(AI)and unmanned aerial vehicles(UAVs)have emerged as potential solutions to develop versatile,dynamic,and energy-efficient communication systems.The study proposes an AI-based UAV architecture that utilizes cooperative reinforcement learning(CoRL)to manage an autonomous network.The UAVs collaborate by sharing local observations and real-time state exchanges to optimize user connectivity,movement directions,allocate power,and resource distribution.Unlike conventional centralized or autonomous methods,CoRL involves joint state sharing and conflict-sensitive reward shaping,which ensures fair coverage,less interference,and enhanced adaptability in a dynamic urban environment.Simulations conducted in smart city scenarios with 10 UAVs and 50 ground users demonstrate that the proposed CoRL-based UAV system increases user coverage by up to 10%,achieves convergence 40%faster,and reduces latency and energy consumption by 30%compared with centralized and decentralized baselines.Furthermore,the distributed nature of the algorithm ensures scalability and flexibility,making it well-suited for future large-scale 6G deployments.The results highlighted that AI-enabled UAV systems enhance connectivity,support ultra-reliable low-latency communications(URLLC),and improve 6G network efficiency.Future work will extend the framework with adaptive modulation,beamforming-aware positioning,and real-world testbed deployment.展开更多
In modern ZnO varistors,traditional aging mechanisms based on increased power consumption are no longer relevant due to reduced power consumption during DC aging.Prolonged exposure to both AC and DC voltages results i...In modern ZnO varistors,traditional aging mechanisms based on increased power consumption are no longer relevant due to reduced power consumption during DC aging.Prolonged exposure to both AC and DC voltages results in increased leakage current,decreased breakdown voltage,and lower nonlinearity,ultimately compromising their protective performance.To investigate the evolution in electrical properties during DC aging,this work developed a finite element model based on Voronoi networks and conducted accelerated aging tests on commercial varistors.Throughout the aging process,current-voltage characteristics and Schottky barrier parameters were measured and analyzed.The results indicate that when subjected to constant voltage,current flows through regions with larger grain sizes,forming discharge channels.As aging progresses,the current focus increases on these channels,leading to a decline in the varistor’s overall performance.Furthermore,analysis of the Schottky barrier parameters shows that the changes in electrical performance during aging are non-monotonic.These findings offer theoretical support for understanding the aging mechanisms and condition assessment of modern stable ZnO varistors.展开更多
基金supported by the National Natural Science Foundation of China under Grant Nos.62076117 and 62166026the Jiangxi Provincial Key Laboratory of Virtual Reality under Grant No.2024SSY03151.
文摘Dynamic sign language recognition holds significant importance, particularly with the application of deep learning to address its complexity. However, existing methods face several challenges. Firstly, recognizing dynamic sign language requires identifying keyframes that best represent the signs, and missing these keyframes reduces accuracy. Secondly, some methods do not focus enough on hand regions, which are small within the overall frame, leading to information loss. To address these challenges, we propose a novel Video Transformer Attention-based Network (VTAN) for dynamic sign language recognition. Our approach prioritizes informative frames and hand regions effectively. To tackle the first issue, we designed a keyframe extraction module enhanced by a convolutional autoencoder, which focuses on selecting information-rich frames and eliminating redundant ones from the video sequences. For the second issue, we developed a soft attention-based transformer module that emphasizes extracting features from hand regions, ensuring that the network pays more attention to hand information within sequences. This dual-focus approach improves effective dynamic sign language recognition by addressing the key challenges of identifying critical frames and emphasizing hand regions. Experimental results on two public benchmark datasets demonstrate the effectiveness of our network, outperforming most of the typical methods in sign language recognition tasks.
基金Key research and development projects of Ningxia,Grant/Award Number:2022BDE03007Natural Science Foundation of Ningxia Province,Grant/Award Numbers:2023A0367,2021A0966,2022AAC05010,2022AAC03004,2021AAC03068。
文摘Network embedding aspires to learn a low-dimensional vector of each node in networks,which can apply to diverse data mining tasks.In real-life,many networks include rich attributes and temporal information.However,most existing embedding approaches ignore either temporal information or network attributes.A self-attention based architecture using higher-order weights and node attributes for both static and temporal attributed network embedding is presented in this article.A random walk sampling algorithm based on higher-order weights and node attributes to capture network topological features is presented.For static attributed networks,the algorithm incorporates first-order to k-order weights,and node attribute similarities into one weighted graph to preserve topological features of networks.For temporal attribute networks,the algorithm incorporates previous snapshots of networks containing first-order to k-order weights,and nodes attribute similarities into one weighted graph.In addition,the algorithm utilises a damping factor to ensure that the more recent snapshots allocate a greater weight.Attribute features are then incorporated into topological features.Next,the authors adopt the most advanced architecture,Self-Attention Networks,to learn node representations.Experimental results on node classification of static attributed networks and link prediction of temporal attributed networks reveal that our proposed approach is competitive against diverse state-of-the-art baseline approaches.
文摘SAR images commonly suffer fromspeckle noise,posing a significant challenge in their analysis and interpretation.Existing convolutional neural network(CNN)based despeckling methods have shown great performance in removing speckle noise.However,these CNN-basedmethods have a fewlimitations.They do not decouple complex background information in amulti-resolutionmanner.Moreover,they have deep network structures thatmay result in many parameters,limiting their applicability tomobile devices.Furthermore,extracting key speckle information in the presence of complex background is also a major problem with SAR.The proposed study addresses these limitations by introducing a lightweight pyramid and attention-based despeckling(PAN-Despeck)network.The primary objective is to enhance image quality and enable improved information interpretation,particularly on mobile devices and scenarios involving complex backgrounds.The PAN-Despeck network leverages domainspecific knowledge and integrates Gaussian Laplacian image pyramid decomposition for multi-resolution image analysis.By utilizing this approach,complex background information can be effectively decoupled,leading to enhanced despeckling performance.Furthermore,the attention mechanism selectively focuses on key speckle features and facilitates complex background removal.The network incorporates recursive and residual blocks to ensure computational efficiency and accelerate training speed,making it lightweight while maintaining high performance.Through comprehensive evaluations,it is demonstrated that PAN-Despeck outperforms existing image restoration methods.With an impressive average peak signal-to-noise ratio(PSNR)of 28.355114 and a remarkable structural similarity index(SSIM)of 0.905467,it demonstrates exceptional performance in effectively reducing speckle noise in SAR images.The source code for the PAN-DeSpeck network is available on GitHub.
基金funded by Naif Arab University for Security Sciences under grant No.NAUSS-23-R11.
文摘In the current digital era,new technologies are becoming an essential part of our lives.Consequently,the number ofmalicious software ormalware attacks is rapidly growing.There is no doubt,themajority ofmalware attacks can be detected by most antivirus programs.However,such types of antivirus programs are one step behind malicious software.Due to these dilemmas,deep learning become popular in the detection and classification of malicious data.Therefore,researchers have significantly focused on finding solutions for malware attacks by analyzing malicious samples with the help of different techniques and models.In this research,we presented a lightweight attention-based novel deep Convolutional Neural Network(DNN-CNN)model for binary and multi-class malware classification,including benign,trojan horse,ransomware,and spyware.We applied the Principal Component Analysis(PCA)technique for feature extraction for binary classification.We used the Synthetic Minority Oversampling Technique(SMOTE)to handle the imbalanced data during multi-class classification.Our proposed attention-based malware detectionmodel is trained on the benchmarkmalware memory dataset named CIC-MalMem-2022.Theresults indicate that our model obtained high accuracy for binary and multi-class classification,99.5% and 97.9%,respectively.
基金The authors gratefully acknowledge support from the NSF CAREER Award DMR 1651668.
文摘In this paper,we demonstrate an application of the Transformer self-attention mechanism in the context of materials science.Our network,the Compositionally Restricted Attention-Based network(CrabNet),explores the area of structure-agnostic materials property predictions when only a chemical formula is provided.Our results show that CrabNet’s performance matches or exceeds current best-practice methods on nearly all of 28 total benchmark datasets.We also demonstrate how CrabNet’s architecture lends itself towards model interpretability by showing different visualization approaches that are made possible by its design.We feel confident that CrabNet and its attention-based framework will be of keen interest to future materials informatics researchers.
基金supported by the Nation Natural Science Foundation of China(NSFC)under Grant No.61462042 and No.61966018.
文摘Traffic flow prediction is an important part of the intelligent transportation system. Accurate multi-step traffic flow prediction plays an important role in improving the operational efficiency of the traffic network. Since traffic flow data has complex spatio-temporal correlation and non-linearity, existing prediction methods are mainly accomplished through a combination of a Graph Convolutional Network (GCN) and a recurrent neural network. The combination strategy has an excellent performance in traffic prediction tasks. However, multi-step prediction error accumulates with the predicted step size. Some scholars use multiple sampling sequences to achieve more accurate prediction results. But it requires high hardware conditions and multiplied training time. Considering the spatiotemporal correlation of traffic flow and influence of external factors, we propose an Attention Based Spatio-Temporal Graph Convolutional Network considering External Factors (ABSTGCN-EF) for multi-step traffic flow prediction. This model models the traffic flow as diffusion on a digraph and extracts the spatial characteristics of traffic flow through GCN. We add meaningful time-slots attention to the encoder-decoder to form an Attention Encoder Network (AEN) to handle temporal correlation. The attention vector is used as a competitive choice to draw the correlation between predicted states and historical states. We considered the impact of three external factors (daytime, weekdays, and traffic accident markers) on the traffic flow prediction tasks. Experiments on two public data sets show that it makes sense to consider external factors. The prediction performance of our ABSTGCN-EF model achieves 7.2%–8.7% higher than the state-of-the-art baselines.
基金the National Natural Science Foundation of China under Grant 62172059 and 62072055Hunan Provincial Natural Science Foundations of China under Grant 2022JJ50318 and 2022JJ30621Scientific Research Fund of Hunan Provincial Education Department of China under Grant 22A0200 and 20K098。
文摘Electrocardiogram(ECG)signal is one of the noninvasive physiological measurement techniques commonly usedin cardiac diagnosis.However,in real scenarios,the ECGsignal is susceptible to various noise erosion,which affectsthe subsequent pathological analysis.Therefore,the effective removal of the noise from ECG signals has becomea top priority in cardiac diagnostic research.Aiming at the problem of incomplete signal shape retention andlow signal-to-noise ratio(SNR)after denoising,a novel ECG denoising network,named attention-based residualdense shrinkage network(ARDSN),is proposed in this paper.Firstly,the shallow ECG characteristics are extractedby a shallow feature extraction network(SFEN).Then,the residual dense shrinkage attention block(RDSAB)isused for adaptive noise suppression.Finally,feature fusion representation(FFR)is performed on the hierarchicalfeatures extracted by a series of RDSABs to reconstruct the de-noised ECG signal.Experiments on the MIT-BIHarrhythmia database and MIT-BIH noise stress test database indicate that the proposed scheme can effectively resistthe interference of different sources of noise on the ECG signal.
文摘A correct and timely fault diagnosis is important for improving the safety and reliability of chemical processes. With the advancement of big data technology, data-driven fault diagnosis methods are being extensively used and still have considerable potential. In recent years, methods based on deep neural networks have made significant breakthroughs, and fault diagnosis methods for industrial processes based on deep learning have attracted considerable research attention. Therefore, we propose a fusion deeplearning algorithm based on a fully convolutional neural network(FCN) to extract features and build models to correctly diagnose all types of faults. We use long short-term memory(LSTM) units to expand our proposed FCN so that our proposed deep learning model can better extract the time-domain features of chemical process data. We also introduce the attention mechanism into the model, aimed at highlighting the importance of features, which is significant for the fault diagnosis of chemical processes with many features. When applied to the benchmark Tennessee Eastman process, our proposed model exhibits impressive performance, demonstrating the effectiveness of the attention-based LSTM FCN in chemical process fault diagnosis.
基金This work was supported in part by the National Natural Science Foundation of China under Grant Nos.61876027,61533020 and 61751312the key T&A program of Chongqing under grant No.cstc2019jscx-mbdxX0048.
文摘It is a common fact that data(features,characteristics or variables)are collected at different sampling frequencies in some fields such as economic and industry.The existing methods usually either ignore the difference from the different sampling frequencies or hardly take notice of the inherent temporal characteristics in mixed frequency data.The authors propose an innovative dual attention-based neural network for mixed frequency data(MID-DualAtt),in order to utilize the inherent temporal characteristics and select the input characteristics reasonably without losing information.According to the authors’knowledge,this is the first study to use the attention mechanism to process mixed fre-quency data.The MID-DualAtt model uses the frequency alignment method to trans-form the high--frequency variables into observation vectors at low frequency,and more critical input characteristics are selected for the current prediction index by attention mechanism.The temporal characteristics are explored by the encoder-decoder with attention based on long-short-term memory networks(LSTM).The proposed MID-DualAtt has been tested in practical application,and the experimental results show that it has better prediction ability than the compared models.
基金The work was supported by the National Key R&D Program of China(Grant No.2020YFC1511601)Fundamental Research Funds for the Central Universities(Grant No.2019SHFWLC01).
文摘Existing almost deep learning methods rely on a large amount of annotated data, so they are inappropriate for forest fire smoke detection with limited data. In this paper, a novel hybrid attention-based few-shot learning method, named Attention-Based Prototypical Network, is proposed for forest fire smoke detection. Specifically, feature extraction network, which consists of convolutional block attention module, could extract high-level and discriminative features and further decrease the false alarm rate resulting from suspected smoke areas. Moreover, we design a metalearning module to alleviate the overfitting issue caused by limited smoke images, and the meta-learning network enables achieving effective detection via comparing the distance between the class prototype of support images and the features of query images. A series of experiments on forest fire smoke datasets and miniImageNet dataset testify that the proposed method is superior to state-of-the-art few-shot learning approaches.
基金National Natural Science Foundation of China,Grant/Award Number:62071039Beijing Natural Science Foundation,Grant/Award Number:L223033。
文摘The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sampling rate,how to model longsequence data and make rational use of the relevant information between channels is also an urgent problem to be solved.In order to solve the above problems,the performance of the end-to-end music separation algorithm is enhanced by improving the network structure.Our main contributions include the following:(1)A more reasonable densely connected U-Net is designed to capture the long-term characteristics of music,such as main melody,tone and so on.(2)On this basis,the multi-head attention and dualpath transformer are introduced in the separation module.Channel attention units are applied recursively on the feature map of each layer of the network,enabling the network to perform long-sequence separation.Experimental results show that after the introduction of the channel attention,the performance of the proposed algorithm has a stable improvement compared with the baseline system.On the MUSDB18 dataset,the average score of the separated audio exceeds that of the current best-performing music separation algorithm based on the time-frequency domain(T-F domain).
基金This work was supported by the National Natural Science Foundation of China(No.61906006).
文摘Whole brain functional connectivity(FC)patterns obtained from resting-state functional magnetic resonance imaging(rs-fMRI)have been widely used in the diagnosis of brain disorders such as autism spectrum disorder(ASD).Recently,an increasing number of studies have focused on employing deep learning techniques to analyze FC patterns for brain disease classification.However,the high dimensionality of the FC features and the interpretation of deep learning results are issues that need to be addressed in the FC-based brain disease classification.In this paper,we proposed a multi-scale attention-based deep neural network(MSA-DNN)model to classify FC patterns for the ASD diagnosis.The model was implemented by adding a flexible multi-scale attention(MSA)module to the auto-encoder based backbone DNN,which can extract multi-scale features of the FC patterns and change the level of attention for different FCs by continuous learning.Our model will reinforce the weights of important FC features while suppress the unimportant FCs to ensure the sparsity of the model weights and enhance the model interpretability.We performed systematic experiments on the large multi-sites ASD dataset with both ten-fold and leaveone-site-out cross-validations.Results showed that our model outperformed classical methods in brain disease classification and revealed robust intersite prediction performance.We also localized important FC features and brain regions associated with ASD classification.Overall,our study further promotes the biomarker detection and computer-aided classification for ASD diagnosis,and the proposed MSA module is flexible and easy to implement in other classification networks.
基金funded by the Hainan Provincial Natural Science Foundation Project(Grant No.622RC675)the National Natural Science Foundation of China(Grant No.62262019).
文摘As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word embedding is utilized to extract information from these templates.However,this method neglects a portion of the content within the logs and confronts the challenge of data imbalance among various log template types after parsing.Currently,specialized research on data imbalance across log template categories remains scarce.A dual-attention-based log anomaly detection model(LogDA),which leveraged data imbalance,was proposed to address these issues in the work.The LogDA model initially utilized a pre-trained model to extract semantic embedding from log templates.Besides,the similarity between embedding was calculated to discern the relationships among the various templates.Then,a Transformer model with a dual-attention mechanism was constructed to capture positional information and global dependencies.Compared to multiple baseline experiments across three public datasets,the proposed approach could improve precision,recall,and F1 scores.
基金This work is supported by the National Natural Science Foundation of China(Grant No.61971057)MoE-CMCC Artifical Intelligence Project(No.MCM20190701).
文摘Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as LSTM and ARIMA are better than convolutional neural network in time series prediction,but they are not enough to mine the periodicity of data.In this article,we perform periodic analysis on two types of time series data,select time metrics with high periodic characteristics,and propose a multi-scale prediction model based on the attention mechanism for the periodic trend of the data.A loss calculation method for traffic time series characteristics is proposed as well.Multiple experiments have been conducted on actual data sets.The experiments show that the method proposed in this paper has better performance than commonly used traffic prediction methods(ARIMA,LSTM,etc.)and 3%-5%increase on MAPE.
基金Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.supported by the National Research Foundation of Korea(NRF)grant funded by the Korea government(*MSIT)(No.2018R1A5A7059549)the Competitive Research Fund of The University of Aizu,Japan.
文摘Real-time surveillance is attributed to recognizing the variety of actions performed by humans.Human Action Recognition(HAR)is a technique that recognizes human actions from a video stream.A range of variations in human actions makes it difficult to recognize with considerable accuracy.This paper presents a novel deep neural network architecture called Attention RB-Net for HAR using video frames.The input is provided to the model in the form of video frames.The proposed deep architecture is based on the unique structuring of residual blocks with several filter sizes.Features are extracted from each frame via several operations with specific parameters defined in the presented novel Attention-based Residual Bottleneck(Attention-RB)DCNN architecture.A fully connected layer receives an attention-based features matrix,and final classification is performed.Several hyperparameters of the proposed model are initialized using Bayesian Optimization(BO)and later utilized in the trained model for testing.In testing,features are extracted from the self-attention layer and passed to neural network classifiers for the final action classification.Two highly cited datasets,HMDB51 and UCF101,were used to validate the proposed architecture and obtained an average accuracy of 87.70%and 97.30%,respectively.The deep convolutional neural network(DCNN)architecture is compared with state-of-the-art(SOTA)methods,including pre-trained models,inside blocks,and recently published techniques,and performs better.
基金sponsored by the National Natural Science Foundation of China(Grant No.52301320)the Natural Science Founds of Fujian Province(No.2023J01790).
文摘Traditional Computational Fluid Dynamics(CFD)simulations are computationally expensive when applied to complex fluid–structure interaction problems and often struggle to capture the essential flow features governing vortex-induced vibrations(VIV)of floating structures.To overcome these limitations,this study develops a hybrid framework that integrates high-fidelity CFD modeling with deep learning techniques to enhance the accuracy and efficiency of VIV response prediction.First,an unstructured finite-volume fluid–structure coupling model is established to generate high-resolution flow field data and extract multi-component time-series feature tensors.These tensors serve as inputs to a Squeeze-and-Excitation Convolutional Neural Network(SE-CNN),which models the nonlinear coupling between flow disturbances and structural responses.The SE-CNN architecture incorporates an attention-based weighting mechanism through an embedded Squeeze-and-Excitation module,dynamically optimizing channel feature importance and improving sensitivity to critical flow characteristics.During training,multidimensional inputs,including pressure,velocity gradient,and displacement sequences,are used to capture the full complexity of fluid–structure interactions.Results demonstrate that the proposed method achieves a maximum amplitude prediction error of only 2.9%and a main frequency deviation below 0.03 Hz,outperforming conventional CNN models by reducing amplitude prediction error from 3.2%to 1.9%.The approach is validated using a representative semi-submersible platform,confirming its robustness across varying damping conditions and flow velocities.
基金the National Natural Science Foundation of China(Grant Nos.61972016 and 62032016)the Beijing Nova Program(Grant No.20220484106)。
文摘We present a novel approach for the prediction of crystal material properties that is distinct from the computationally complex and expensive density functional theory(DFT)-based calculations.Instead,we utilize an attention-based graph neural network that yields high-accuracy predictions.Our approach employs two attention mechanisms that allow for message passing on the crystal graphs,which in turn enable the model to selectively attend to pertinent atoms and their local environments,thereby improving performance.We conduct comprehensive experiments to validate our approach,which demonstrates that our method surpasses existing methods in terms of predictive accuracy.Our results suggest that deep learning,particularly attention-based networks,holds significant promise for predicting crystal material properties,with implications for material discovery and the refined intelligent systems.
基金supported by the Chung-Ang University Research Grants in 2023.Alsothe work is supported by the ELLIIT Excellence Center at Linköping–Lund in Information Technology in Sweden.
文摘Recommending personalized travel routes from sparse,implicit feedback poses a significant challenge,as conventional systems often struggle with information overload and fail to capture the complex,sequential nature of user preferences.To address this,we propose a Conditional Generative Adversarial Network(CGAN)that generates diverse and highly relevant itineraries.Our approach begins by constructing a conditional vector that encapsulates a user’s profile.This vector uniquely fuses embeddings from a Heterogeneous Information Network(HIN)to model complex user-place-route relationships,a Recurrent Neural Network(RNN)to capture sequential path dynamics,and Neural Collaborative Filtering(NCF)to incorporate collaborative signals from the wider user base.This comprehensive condition,further enhanced with features representing user interaction confidence and uncertainty,steers a CGAN stabilized by spectral normalization to generate high-fidelity latent route representations,effectively mitigating the data sparsity problem.Recommendations are then formulated using an Anchor-and-Expand algorithm,which selects relevant starting Points of Interest(POI)based on user history,then expands routes through latent similarity matching and geographic coherence optimization,culminating in Traveling Salesman Problem(TSP)-based route optimization for practical travel distances.Experiments on a real-world check-in dataset validate our model’s unique generative capability,achieving F1 scores ranging from 0.163 to 0.305,and near-zero pairs−F1 scores between 0.002 and 0.022.These results confirm the model’s success in generating novel travel routes by recommending new locations and sequences rather than replicating users’past itineraries.This work provides a robust solution for personalized travel planning,capable of generating novel and compelling routes for both new and existing users by learning from collective travel intelligence.
基金supported by the National Research Foundation of Korea(NRF)grant funded by the Korea government(MSIT)(RS-2025-00559546)supported by the IITP(Institute of Information&Coummunications Technology Planning&Evaluation)-ITRC(Information Technology Research Center)grant funded by the Korea government(Ministry of Science and ICT)(IITP-2025-RS-2023-00259004).
文摘The advent of sixth-generation(6G)networks introduces unprecedented challenges in achieving seamless connectivity,ultra-low latency,and efficient resource management in highly dynamic environments.Although fifth-generation(5G)networks transformed mobile broadband and machine-type communications at massive scales,their properties of scaling,interference management,and latency remain a limitation in dense high mobility settings.To overcome these limitations,artificial intelligence(AI)and unmanned aerial vehicles(UAVs)have emerged as potential solutions to develop versatile,dynamic,and energy-efficient communication systems.The study proposes an AI-based UAV architecture that utilizes cooperative reinforcement learning(CoRL)to manage an autonomous network.The UAVs collaborate by sharing local observations and real-time state exchanges to optimize user connectivity,movement directions,allocate power,and resource distribution.Unlike conventional centralized or autonomous methods,CoRL involves joint state sharing and conflict-sensitive reward shaping,which ensures fair coverage,less interference,and enhanced adaptability in a dynamic urban environment.Simulations conducted in smart city scenarios with 10 UAVs and 50 ground users demonstrate that the proposed CoRL-based UAV system increases user coverage by up to 10%,achieves convergence 40%faster,and reduces latency and energy consumption by 30%compared with centralized and decentralized baselines.Furthermore,the distributed nature of the algorithm ensures scalability and flexibility,making it well-suited for future large-scale 6G deployments.The results highlighted that AI-enabled UAV systems enhance connectivity,support ultra-reliable low-latency communications(URLLC),and improve 6G network efficiency.Future work will extend the framework with adaptive modulation,beamforming-aware positioning,and real-world testbed deployment.
文摘In modern ZnO varistors,traditional aging mechanisms based on increased power consumption are no longer relevant due to reduced power consumption during DC aging.Prolonged exposure to both AC and DC voltages results in increased leakage current,decreased breakdown voltage,and lower nonlinearity,ultimately compromising their protective performance.To investigate the evolution in electrical properties during DC aging,this work developed a finite element model based on Voronoi networks and conducted accelerated aging tests on commercial varistors.Throughout the aging process,current-voltage characteristics and Schottky barrier parameters were measured and analyzed.The results indicate that when subjected to constant voltage,current flows through regions with larger grain sizes,forming discharge channels.As aging progresses,the current focus increases on these channels,leading to a decline in the varistor’s overall performance.Furthermore,analysis of the Schottky barrier parameters shows that the changes in electrical performance during aging are non-monotonic.These findings offer theoretical support for understanding the aging mechanisms and condition assessment of modern stable ZnO varistors.
