Pneumonia is a serious disease that can be fatal,particularly among children and the elderly.The accuracy of pneumonia diagnosis can be improved by combining artificial-intelligence technology with X-ray imaging.This ...Pneumonia is a serious disease that can be fatal,particularly among children and the elderly.The accuracy of pneumonia diagnosis can be improved by combining artificial-intelligence technology with X-ray imaging.This study proposes X-ODFCANet,which addresses the issues of low accuracy and excessive parameters in existing deep-learningbased pneumonia-classification methods.This network incorporates a feature coordination attention module and an omni-dimensional dynamic convolution(ODConv)module,leveraging the residual module for feature extraction from X-ray images.The feature coordination attention module utilizes two one-dimensional feature encoding processes to aggregate feature information from different spatial directions.Additionally,the ODConv module extracts and fuses feature information in four dimensions:the spatial dimension of the convolution kernel,input and output channel quantities,and convolution kernel quantity.The experimental results demonstrate that the proposed method can effectively improve the accuracy of pneumonia classification,which is 3.77%higher than that of ResNet18.The model parameters are 4.45M,which was reduced by approximately 2.5 times.The code is available at https://github.com/limuni/X ODFCA NET.展开更多
For the existing deep learning image restoration methods,the joint guidance of structure and texture information is not considered,which leads to structural disorder and texture blur in the restoration results.A gener...For the existing deep learning image restoration methods,the joint guidance of structure and texture information is not considered,which leads to structural disorder and texture blur in the restoration results.A generative adversarial mural inpainting algorithm based on structural and texture hybrid enhancement was proposed.Firstly,the structure guidance branch composed of dynamic convolution cascade was constructed to improve the expression ability of structure features,and the structure information was used to guide the encoder coding to enhance the edge contour information of the coding feature map.Then,the multi-granularity feature extraction module was designed to obtain the texture features of texture guided branches,and the multi-scale texture information was used to guide the decoder to reconstruct and repair,so as to improve the texture consistency of murals.Finally,skip connection was used to promote the feature sharing of structure and texture features,and the spectral-normalized PatchGAN discriminator was used to complete the mural restoration.The digital restoration experiment results of real Dunhuang murals showed that the proposed method was better than the comparison algorithms in both subjective and objective evaluation,and the restoration results were clearer and more natural.展开更多
In the production processes of modern industry,accurate assessment of the system’s health state and traceability non-optimal factors are key to ensuring“safe,stable,long-term,full load and optimal”operation of the ...In the production processes of modern industry,accurate assessment of the system’s health state and traceability non-optimal factors are key to ensuring“safe,stable,long-term,full load and optimal”operation of the production process.The benzene-to-ethylene ratio control system is a complex system based on anMPC-PID doublelayer architecture.Taking into consideration the interaction between levels,coupling between loops and conditions of incomplete operation data,this paper proposes a health assessment method for the dual-layer control system by comprehensively utilizing deep learning technology.Firstly,according to the results of the pre-assessment of the system layers and loops bymultivariate statisticalmethods,seven characteristic parameters that have a significant impact on the health state of the system are identified.Next,aiming at the problem of incomplete assessment data set due to the uneven distribution of actual system operating health state,the original unbalanced dataset is augmented using aWasserstein generative adversarial network with gradient penalty term,and a complete dataset is obtained to characterise all the health states of the system.On this basis,a new deep learning-based health assessment framework for the benzeneto-ethylene ratio control system is constructed based on traditionalmultivariate statistical assessment.This framework can overcome the shortcomings of the linear weighted fusion related to the coupling and nonlinearity of the subsystem health state at different layers,and reduce the dependence of the prior knowledge.Furthermore,by introducing a dynamic attention mechanism(AM)into the convolutional neural network(CNN),the assessment model integrating both assessment and traceability is constructed,which can achieve the health assessment and trace the non-optimal factors of the complex control systems with the double-layer architecture.Finally,the effectiveness and superiority of the proposed method have been verified by the benzene-ethylene ratio control system of the alkylation process unit in a styrene plant.展开更多
Latent information is difficult to get from the text in speech synthesis.Studies show that features from speech can get more information to help text encoding.In the field of speech encoding,a lot of work has been con...Latent information is difficult to get from the text in speech synthesis.Studies show that features from speech can get more information to help text encoding.In the field of speech encoding,a lot of work has been conducted on two aspects.The first aspect is to encode speech frame by frame.The second aspect is to encode the whole speech to a vector.But the scale in these aspects is fixed.So,encoding speech with an adjustable scale for more latent information is worthy of investigation.But current alignment approaches only support frame-by-frame encoding and speech-to-vector encoding.It remains a challenge to propose a new alignment approach to support adjustable scale speech encoding.This paper presents the dynamic speech encoder with a new alignment approach in conjunction with frame-by-frame encoding and speech-to-vector encoding.The speech feature fromourmodel achieves three functions.First,the speech feature can reconstruct the origin speech while the length of the speech feature is equal to the text length.Second,our model can get text embedding fromspeech,and the encoded speech feature is similar to the text embedding result.Finally,it can transfer the style of synthesis speech and make it more similar to the given reference speech.展开更多
Due to self-occlusion and high degree of freedom,estimating 3D hand pose from a single RGB image is a great challenging problem.Graph convolutional networks(GCNs)use graphs to describe the physical connection relation...Due to self-occlusion and high degree of freedom,estimating 3D hand pose from a single RGB image is a great challenging problem.Graph convolutional networks(GCNs)use graphs to describe the physical connection relationships between hand joints and improve the accuracy of 3D hand pose regression.However,GCNs cannot effectively describe the relationships between non-adjacent hand joints.Recently,hypergraph convolutional networks(HGCNs)have received much attention as they can describe multi-dimensional relationships between nodes through hyperedges;therefore,this paper proposes a framework for 3D hand pose estimation based on HGCN,which can better extract correlated relationships between adjacent and non-adjacent hand joints.To overcome the shortcomings of predefined hypergraph structures,a kind of dynamic hypergraph convolutional network is proposed,in which hyperedges are constructed dynamically based on hand joint feature similarity.To better explore the local semantic relationships between nodes,a kind of semantic dynamic hypergraph convolution is proposed.The proposed method is evaluated on publicly available benchmark datasets.Qualitative and quantitative experimental results both show that the proposed HGCN and improved methods for 3D hand pose estimation are better than GCN,and achieve state-of-the-art performance compared with existing methods.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
At present,super-resolution algorithms are employed to tackle the challenge of low image resolution,but it is difficult to extract differentiated feature details based on various inputs,resulting in poor generalizatio...At present,super-resolution algorithms are employed to tackle the challenge of low image resolution,but it is difficult to extract differentiated feature details based on various inputs,resulting in poor generalization ability.Given this situation,this study first analyzes the features of some feature extraction modules of the current super-resolution algorithm and then proposes an adaptive feature fusion block(AFB)for feature extraction.This module mainly comprises dynamic convolution,attention mechanism,and pixel-based gating mechanism.Combined with dynamic convolution with scale information,the network can extract more differentiated feature information.The introduction of a channel spatial attention mechanism combined with multi-feature fusion further enables the network to retain more important feature information.Dynamic convolution and pixel-based gating mechanisms enhance the module’s adaptability.Finally,a comparative experiment of a super-resolution algorithm based on the AFB module is designed to substantiate the efficiency of the AFB module.The results revealed that the network combined with the AFB module has stronger generalization ability and expression ability.展开更多
Accurately predicting the chiller coefficient of performance(COP)is essential for improving the energy efficiency of heating,ventilation,and air conditioning(HVAC)systems,significantly contributing to energy conservat...Accurately predicting the chiller coefficient of performance(COP)is essential for improving the energy efficiency of heating,ventilation,and air conditioning(HVAC)systems,significantly contributing to energy conservation in buildings.Traditional performance prediction methods often overlook the dynamic interaction among sensor variables and face challenges in using extensive historical data efficiently,which impedes accurate predictions.To overcome these challenges,this paper proposes an innovative on-site chiller performance prediction method employing a dynamic graph convolutional network(GCN)enhanced by association rules.The distinctive feature of this method is constructing an association graph bank containing static graphs in each operating mode by mining the association rules between various sensor variables in historical operating data.A real-time graph is created by analyzing the correlation between various sensor variables in the current operating data.This graph is fused online with the static graph in the current operating mode to obtain a dynamic graph used for feature extraction and training of GCN.The effectiveness of this method has been empirically confirmed through the operational data of an actual building chiller system.Comparative analysis with state-of-the-art methods highlights the superior performance of the proposed method.展开更多
基金supported in part by the Key Research and Development Program of Shaanxi Province of China,No.2024GX-YBXM-149in part by the National Natural Science Foundation of China,No.62071381.
文摘Pneumonia is a serious disease that can be fatal,particularly among children and the elderly.The accuracy of pneumonia diagnosis can be improved by combining artificial-intelligence technology with X-ray imaging.This study proposes X-ODFCANet,which addresses the issues of low accuracy and excessive parameters in existing deep-learningbased pneumonia-classification methods.This network incorporates a feature coordination attention module and an omni-dimensional dynamic convolution(ODConv)module,leveraging the residual module for feature extraction from X-ray images.The feature coordination attention module utilizes two one-dimensional feature encoding processes to aggregate feature information from different spatial directions.Additionally,the ODConv module extracts and fuses feature information in four dimensions:the spatial dimension of the convolution kernel,input and output channel quantities,and convolution kernel quantity.The experimental results demonstrate that the proposed method can effectively improve the accuracy of pneumonia classification,which is 3.77%higher than that of ResNet18.The model parameters are 4.45M,which was reduced by approximately 2.5 times.The code is available at https://github.com/limuni/X ODFCA NET.
基金supported by Ministry of Education in China Project of Humanities and Social Sciences(No.19YJC760012)Star of Innovation Project for Outstanding Graduate Students in Gansu Province(No.2022CXZX-546)。
文摘For the existing deep learning image restoration methods,the joint guidance of structure and texture information is not considered,which leads to structural disorder and texture blur in the restoration results.A generative adversarial mural inpainting algorithm based on structural and texture hybrid enhancement was proposed.Firstly,the structure guidance branch composed of dynamic convolution cascade was constructed to improve the expression ability of structure features,and the structure information was used to guide the encoder coding to enhance the edge contour information of the coding feature map.Then,the multi-granularity feature extraction module was designed to obtain the texture features of texture guided branches,and the multi-scale texture information was used to guide the decoder to reconstruct and repair,so as to improve the texture consistency of murals.Finally,skip connection was used to promote the feature sharing of structure and texture features,and the spectral-normalized PatchGAN discriminator was used to complete the mural restoration.The digital restoration experiment results of real Dunhuang murals showed that the proposed method was better than the comparison algorithms in both subjective and objective evaluation,and the restoration results were clearer and more natural.
基金supported by the National Science Foundation of China(62263020)the Key Project of Natural Science Foundation of Gansu Province(25JRRA061)+1 种基金the Key R&D Program of Gansu Province(23YFGA0061)the Scientific Research Initiation Fund of Lanzhou University of Technology(061602).
文摘In the production processes of modern industry,accurate assessment of the system’s health state and traceability non-optimal factors are key to ensuring“safe,stable,long-term,full load and optimal”operation of the production process.The benzene-to-ethylene ratio control system is a complex system based on anMPC-PID doublelayer architecture.Taking into consideration the interaction between levels,coupling between loops and conditions of incomplete operation data,this paper proposes a health assessment method for the dual-layer control system by comprehensively utilizing deep learning technology.Firstly,according to the results of the pre-assessment of the system layers and loops bymultivariate statisticalmethods,seven characteristic parameters that have a significant impact on the health state of the system are identified.Next,aiming at the problem of incomplete assessment data set due to the uneven distribution of actual system operating health state,the original unbalanced dataset is augmented using aWasserstein generative adversarial network with gradient penalty term,and a complete dataset is obtained to characterise all the health states of the system.On this basis,a new deep learning-based health assessment framework for the benzeneto-ethylene ratio control system is constructed based on traditionalmultivariate statistical assessment.This framework can overcome the shortcomings of the linear weighted fusion related to the coupling and nonlinearity of the subsystem health state at different layers,and reduce the dependence of the prior knowledge.Furthermore,by introducing a dynamic attention mechanism(AM)into the convolutional neural network(CNN),the assessment model integrating both assessment and traceability is constructed,which can achieve the health assessment and trace the non-optimal factors of the complex control systems with the double-layer architecture.Finally,the effectiveness and superiority of the proposed method have been verified by the benzene-ethylene ratio control system of the alkylation process unit in a styrene plant.
基金supported by National Key R&D Program of China (2020AAA0107901).
文摘Latent information is difficult to get from the text in speech synthesis.Studies show that features from speech can get more information to help text encoding.In the field of speech encoding,a lot of work has been conducted on two aspects.The first aspect is to encode speech frame by frame.The second aspect is to encode the whole speech to a vector.But the scale in these aspects is fixed.So,encoding speech with an adjustable scale for more latent information is worthy of investigation.But current alignment approaches only support frame-by-frame encoding and speech-to-vector encoding.It remains a challenge to propose a new alignment approach to support adjustable scale speech encoding.This paper presents the dynamic speech encoder with a new alignment approach in conjunction with frame-by-frame encoding and speech-to-vector encoding.The speech feature fromourmodel achieves three functions.First,the speech feature can reconstruct the origin speech while the length of the speech feature is equal to the text length.Second,our model can get text embedding fromspeech,and the encoded speech feature is similar to the text embedding result.Finally,it can transfer the style of synthesis speech and make it more similar to the given reference speech.
基金the National Key Research and Development Program of China(No.2021ZD0111902)the National Natural Science Foundation of China(Nos.62172022 and U21B2038)。
文摘Due to self-occlusion and high degree of freedom,estimating 3D hand pose from a single RGB image is a great challenging problem.Graph convolutional networks(GCNs)use graphs to describe the physical connection relationships between hand joints and improve the accuracy of 3D hand pose regression.However,GCNs cannot effectively describe the relationships between non-adjacent hand joints.Recently,hypergraph convolutional networks(HGCNs)have received much attention as they can describe multi-dimensional relationships between nodes through hyperedges;therefore,this paper proposes a framework for 3D hand pose estimation based on HGCN,which can better extract correlated relationships between adjacent and non-adjacent hand joints.To overcome the shortcomings of predefined hypergraph structures,a kind of dynamic hypergraph convolutional network is proposed,in which hyperedges are constructed dynamically based on hand joint feature similarity.To better explore the local semantic relationships between nodes,a kind of semantic dynamic hypergraph convolution is proposed.The proposed method is evaluated on publicly available benchmark datasets.Qualitative and quantitative experimental results both show that the proposed HGCN and improved methods for 3D hand pose estimation are better than GCN,and achieve state-of-the-art performance compared with existing methods.
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 Sichuan Science and Technology Program(2021YFQ0003,2023YFSY0026,2023YFH0004).
文摘At present,super-resolution algorithms are employed to tackle the challenge of low image resolution,but it is difficult to extract differentiated feature details based on various inputs,resulting in poor generalization ability.Given this situation,this study first analyzes the features of some feature extraction modules of the current super-resolution algorithm and then proposes an adaptive feature fusion block(AFB)for feature extraction.This module mainly comprises dynamic convolution,attention mechanism,and pixel-based gating mechanism.Combined with dynamic convolution with scale information,the network can extract more differentiated feature information.The introduction of a channel spatial attention mechanism combined with multi-feature fusion further enables the network to retain more important feature information.Dynamic convolution and pixel-based gating mechanisms enhance the module’s adaptability.Finally,a comparative experiment of a super-resolution algorithm based on the AFB module is designed to substantiate the efficiency of the AFB module.The results revealed that the network combined with the AFB module has stronger generalization ability and expression ability.
基金supported in part by the Science and Technology Innovation Program of Hunan Province(No.2022RC1090)in part by the National Natural Science Foundation of China(No.62173349)+2 种基金in part by the Natural Science Foundation of Hunan Province(No.2022J20076)in part by the Innovation Driven Projection of Central South University(No.2023CXQD073)in part by the Major Program of Xiangjiang Laboratory(No.22XJ01005).
文摘Accurately predicting the chiller coefficient of performance(COP)is essential for improving the energy efficiency of heating,ventilation,and air conditioning(HVAC)systems,significantly contributing to energy conservation in buildings.Traditional performance prediction methods often overlook the dynamic interaction among sensor variables and face challenges in using extensive historical data efficiently,which impedes accurate predictions.To overcome these challenges,this paper proposes an innovative on-site chiller performance prediction method employing a dynamic graph convolutional network(GCN)enhanced by association rules.The distinctive feature of this method is constructing an association graph bank containing static graphs in each operating mode by mining the association rules between various sensor variables in historical operating data.A real-time graph is created by analyzing the correlation between various sensor variables in the current operating data.This graph is fused online with the static graph in the current operating mode to obtain a dynamic graph used for feature extraction and training of GCN.The effectiveness of this method has been empirically confirmed through the operational data of an actual building chiller system.Comparative analysis with state-of-the-art methods highlights the superior performance of the proposed method.
