Massive multiple-input multiple-output(MIMO)technology enables higher data rate transmission in the future mobile communications.However,exploiting a large number of antenna elements at base station(BS)makes effective...Massive multiple-input multiple-output(MIMO)technology enables higher data rate transmission in the future mobile communications.However,exploiting a large number of antenna elements at base station(BS)makes effective implementation of massive MIMO challenging,due to the size and weight limits of the masssive MIMO that are located on each BS.Therefore,in order to miniaturize the massive MIMO,it is crucial to reduce the number of antenna elements via effective methods such as sparse array synthesis.In this paper,a multiple-pattern synthesis is considered towards convex optimization(CO).The joint convex optimization(JCO)based synthesis is proposed to construct a codebook for beamforming.Then,a criterion containing multiple constraints is developed,in which the sparse array is required to fullfill all constraints.Finally,extensive evaluations are performed under realistic simulation settings.The results show that with the same number of antenna elements,sparse array using the proposed JCO-based synthesis outperforms not only the uniform array,but also the sparse array with the existing CO-based synthesis method.Furthermore,with a half of the number of antenna elements that on the uniform array,the performance of the JCO-based sparse array approaches to that of the uniform array.展开更多
针对噪声影响工件图像分割、跟踪等问题,给出一种基于自适应平方变换方法.首先将噪声图像中提取的噪声图像块减去块均值,固定稀疏水平,学习平方变换,更新稀疏水平,作为下一次学习平方变换的稀疏水平,然后更新迭代学习平方变换和稀疏水平...针对噪声影响工件图像分割、跟踪等问题,给出一种基于自适应平方变换方法.首先将噪声图像中提取的噪声图像块减去块均值,固定稀疏水平,学习平方变换,更新稀疏水平,作为下一次学习平方变换的稀疏水平,然后更新迭代学习平方变换和稀疏水平,最后一次迭代的去噪块的均值估计作为去噪图像.实验结果表明,给出的方法能较好地滤除噪声.与核奇值分解(K-SVD,kernal singular value decompostion)算法相比,该算法去噪后图像的峰值信噪比(PSNR,peak signal to noise ratio)约是K-SVD算法的2倍,去噪速度是K-SVD的3.9倍.展开更多
With a goal to optimize the element positions to reduce the peak sidelobe level （PSLL） of the array pattern, a modified real Genetic Algorithms （MGA） for the synthesis of sparse linear arrays is described. The mul...With a goal to optimize the element positions to reduce the peak sidelobe level （PSLL） of the array pattern, a modified real Genetic Algorithms （MGA） for the synthesis of sparse linear arrays is described. The multiple optimization constrains include the number of elements, the aperture and the minimum element spacing. The advanced new approach reduces the size of the searching area of GA by means of indirect description of chromosome and avoids infeasible solution during the optimization process by designing the new genetic operators. The elementary steps of MGA are presented. The simulated results confirm the great efficiency and the robustness of this algorithm.展开更多
Stance detection is the task of attitude identification toward a standpoint.Previous work of stance detection has focused on feature extraction but ignored the fact that irrelevant features exist as noise during highe...Stance detection is the task of attitude identification toward a standpoint.Previous work of stance detection has focused on feature extraction but ignored the fact that irrelevant features exist as noise during higher-level abstracting.Moreover,because the target is not always mentioned in the text,most methods have ignored target information.In order to solve these problems,we propose a neural network ensemble method that combines the timing dependence bases on long short-term memory(LSTM)and the excellent extracting performance of convolutional neural networks(CNNs).The method can obtain multi-level features that consider both local and global features.We also introduce attention mechanisms to magnify target information-related features.Furthermore,we employ sparse coding to remove noise to obtain characteristic features.Performance was improved by using sparse coding on the basis of attention employment and feature extraction.We evaluate our approach on the SemEval-2016Task 6-A public dataset,achieving a performance that exceeds the benchmark and those of participating teams.展开更多
Purpose-The real-time generation of feature descriptors for object recognition is a challenging problem.In this research,the purpose of this paper is to provide a hardware friendly framework to generate sparse feature...Purpose-The real-time generation of feature descriptors for object recognition is a challenging problem.In this research,the purpose of this paper is to provide a hardware friendly framework to generate sparse features that can be useful for key feature point selection,feature extraction,and descriptor construction.The inspiration is drawn from feature formation processes of the human brain,taking into account the sparse,modular,and hierarchical processing of visual information.Design/methodology/approach-A sparse set of neurons referred as active neurons determines the feature points necessary for high-level vision applications such as object recognition.A psycho-physical mechanism of human low-level vision relates edge detection to noticeable local spatial stimuli,representing this set of active neurons.A cognitive memory cell array-based implementation of low-level vision is proposed.Applications of memory cell in edge detection are used for realizing human vision inspired feature selection and leading to feature vector construction for high-level vision applications.Findings-True parallel architecture and faster response of cognitive circuits avoid time costly and redundant feature extraction steps.Validation of proposed feature vector toward high-level computer vision applications is demonstrated using standard object recognition databases.The comparison against existing state-of-the-art object recognition features and methods shows an accuracy of 97,95,69 percent for Columbia Object Image Library-100,ALOI,and PASCAL VOC 2007 databases indicating an increase from benchmark methods by 5,3 and 10 percent,respectively.Originality/value-A hardware friendly low-level sparse edge feature processing system isproposed for recognizing objects.The edge features are developed based on threshold logic of neurons,and the sparse selection of the features applies a modular and hierarchical processing inspired from the human neural system.展开更多
Based on the recently developed data-driven time-frequency analysis(Hou and Shi, 2013), we propose a two-level method to look for the sparse time-frequency decomposition of multiscale data. In the two-level method, we...Based on the recently developed data-driven time-frequency analysis(Hou and Shi, 2013), we propose a two-level method to look for the sparse time-frequency decomposition of multiscale data. In the two-level method, we first run a local algorithm to get a good approximation of the instantaneous frequency. We then pass this instantaneous frequency to the global algorithm to get an accurate global intrinsic mode function(IMF)and instantaneous frequency. The two-level method alleviates the difficulty of the mode mixing to some extent.We also present a method to reduce the end effects.展开更多
文摘Massive multiple-input multiple-output(MIMO)technology enables higher data rate transmission in the future mobile communications.However,exploiting a large number of antenna elements at base station(BS)makes effective implementation of massive MIMO challenging,due to the size and weight limits of the masssive MIMO that are located on each BS.Therefore,in order to miniaturize the massive MIMO,it is crucial to reduce the number of antenna elements via effective methods such as sparse array synthesis.In this paper,a multiple-pattern synthesis is considered towards convex optimization(CO).The joint convex optimization(JCO)based synthesis is proposed to construct a codebook for beamforming.Then,a criterion containing multiple constraints is developed,in which the sparse array is required to fullfill all constraints.Finally,extensive evaluations are performed under realistic simulation settings.The results show that with the same number of antenna elements,sparse array using the proposed JCO-based synthesis outperforms not only the uniform array,but also the sparse array with the existing CO-based synthesis method.Furthermore,with a half of the number of antenna elements that on the uniform array,the performance of the JCO-based sparse array approaches to that of the uniform array.
文摘针对噪声影响工件图像分割、跟踪等问题,给出一种基于自适应平方变换方法.首先将噪声图像中提取的噪声图像块减去块均值,固定稀疏水平,学习平方变换,更新稀疏水平,作为下一次学习平方变换的稀疏水平,然后更新迭代学习平方变换和稀疏水平,最后一次迭代的去噪块的均值估计作为去噪图像.实验结果表明,给出的方法能较好地滤除噪声.与核奇值分解(K-SVD,kernal singular value decompostion)算法相比,该算法去噪后图像的峰值信噪比(PSNR,peak signal to noise ratio)约是K-SVD算法的2倍,去噪速度是K-SVD的3.9倍.
基金Supported by National Defense Science and Technology Key Laboratory Foundation Project of China
文摘With a goal to optimize the element positions to reduce the peak sidelobe level （PSLL） of the array pattern, a modified real Genetic Algorithms （MGA） for the synthesis of sparse linear arrays is described. The multiple optimization constrains include the number of elements, the aperture and the minimum element spacing. The advanced new approach reduces the size of the searching area of GA by means of indirect description of chromosome and avoids infeasible solution during the optimization process by designing the new genetic operators. The elementary steps of MGA are presented. The simulated results confirm the great efficiency and the robustness of this algorithm.
基金This work is supported by the Fundamental Research Funds for the Central Universities(Grant No.2572019BH03).
文摘Stance detection is the task of attitude identification toward a standpoint.Previous work of stance detection has focused on feature extraction but ignored the fact that irrelevant features exist as noise during higher-level abstracting.Moreover,because the target is not always mentioned in the text,most methods have ignored target information.In order to solve these problems,we propose a neural network ensemble method that combines the timing dependence bases on long short-term memory(LSTM)and the excellent extracting performance of convolutional neural networks(CNNs).The method can obtain multi-level features that consider both local and global features.We also introduce attention mechanisms to magnify target information-related features.Furthermore,we employ sparse coding to remove noise to obtain characteristic features.Performance was improved by using sparse coding on the basis of attention employment and feature extraction.We evaluate our approach on the SemEval-2016Task 6-A public dataset,achieving a performance that exceeds the benchmark and those of participating teams.
文摘Purpose-The real-time generation of feature descriptors for object recognition is a challenging problem.In this research,the purpose of this paper is to provide a hardware friendly framework to generate sparse features that can be useful for key feature point selection,feature extraction,and descriptor construction.The inspiration is drawn from feature formation processes of the human brain,taking into account the sparse,modular,and hierarchical processing of visual information.Design/methodology/approach-A sparse set of neurons referred as active neurons determines the feature points necessary for high-level vision applications such as object recognition.A psycho-physical mechanism of human low-level vision relates edge detection to noticeable local spatial stimuli,representing this set of active neurons.A cognitive memory cell array-based implementation of low-level vision is proposed.Applications of memory cell in edge detection are used for realizing human vision inspired feature selection and leading to feature vector construction for high-level vision applications.Findings-True parallel architecture and faster response of cognitive circuits avoid time costly and redundant feature extraction steps.Validation of proposed feature vector toward high-level computer vision applications is demonstrated using standard object recognition databases.The comparison against existing state-of-the-art object recognition features and methods shows an accuracy of 97,95,69 percent for Columbia Object Image Library-100,ALOI,and PASCAL VOC 2007 databases indicating an increase from benchmark methods by 5,3 and 10 percent,respectively.Originality/value-A hardware friendly low-level sparse edge feature processing system isproposed for recognizing objects.The edge features are developed based on threshold logic of neurons,and the sparse selection of the features applies a modular and hierarchical processing inspired from the human neural system.
基金supported by National Science Foundation of USA (Grants Nos. DMS1318377 and DMS-1613861)National Natural Science Foundation of China (Grant Nos. 11371220, 11671005, 11371173, 11301222 and 11526096)
文摘Based on the recently developed data-driven time-frequency analysis(Hou and Shi, 2013), we propose a two-level method to look for the sparse time-frequency decomposition of multiscale data. In the two-level method, we first run a local algorithm to get a good approximation of the instantaneous frequency. We then pass this instantaneous frequency to the global algorithm to get an accurate global intrinsic mode function(IMF)and instantaneous frequency. The two-level method alleviates the difficulty of the mode mixing to some extent.We also present a method to reduce the end effects.
