检索结果-内蒙古大学图书馆

Generalized Modified Blake-Zisserman Robust Sparse Adaptive Filters

IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS 2023年第1期53卷 647-652页

作者： Kumar, Krishna Karthik, M. L. N. S. George, Nithin V. Indian Inst Technol Gandhinagar Dept Elect Engn Gandhinagar 382355 India

In the past years, the generalized maximum correntropy criterion (GMCC) has been widely used in adaptive filters to provide robust behavior under non-Gaussian/impulsive noise environments. However, GMCC-based adaptive filters are affected by high steady-state misalignment. In order to enhance the robustness under non-Gaussian noise environments and reduce steady-state misalignment, a generalized modified Blake-Zisserman (GMBZ) robust loss function is introduced in this correspondence. Furthermore, a GMBZ adaptive filter (GMBZ-AF) has been developed that provides improved convergence performance over other existing algorithms. The proposed learning scheme has a computational complexity very similar to that of the GMCC-based adaptive filtering method. In order to further exploit the sparse nature of the system for identifying sparse systems and simultaneously provide robust convergence, two new robust sparse adaptive filters: 1) zero attracting GMBZ-AF (ZA-GMBZ-AF) and 2) reweighted ZA-GMBZ-AF (RZA-GMBZ-AF) have also been proposed. To further enhance the filter convergence performance, a new robust and sparsity-aware loss function called generalized modified dual Blake-Zisserman (GMDBZ) is also introduced in this correspondence and the corresponding GMDBZ adaptive filter (GMDBZ-AF) has been developed.

关键词： Cost function signal processing algorithms Steady-state Convergence Computational complexity Background noise Adaptation models Adaptive filters impulsive noise maximum correntropy criterion (MCC) robust learning sparse system identification

来源：评论

学校读者我要写书评

暂无评论

Fast algorithms for Deep Octonion Networks

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2023年第1期34卷 543-548页

作者： Cariow, Aleksandr Cariowa, Galina West Pomeranian Univ Technol Dept Comp Sci & Informat Technol PL-70310 Szczecin Poland West Pomeranian Univ Technol Dept Comp Sci & Technol PL-70310 Szczecin Poland

This brief presents the results of a study of the possibilities of reducing the arithmetic complexity of computing basic operations in octonionic neural networks and also proposes new algorithmic solutions for efficiently performing these operations. Here, we primarily mean the operation of multiplying octonions, the operation of computing the dot product of two octonion-valued vectors, and the operation of multiple multiplications of an octonion by several other octonions. In order to reduce the computational complexity of these operations, it is proposed to use the fast Walsh-Hadamard transform, which is well known in digital signal processing. Using this transform reduces the number of multiplications and additions of real numbers required to perform computations. Thus, the use of the proposed algorithms will speed up computations in octonion-valued neural networks.

关键词： Neural networks Transforms signal processing algorithms Learning systems Symmetric matrices Sparse matrices Deep learning Acceleration of calculations deep neural networks fast Walsh-Hadamard transform (FWHT) octonions

来源：评论

学校读者我要写书评

暂无评论

Nonlinear Spline Versoria Prioritization Optimization Adaptive Filter for Alpha-Stable Clutter

引用

IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS 2023年第1期59卷 734-744页

作者： Guo, Wenyan Zhi, Yongfeng Northwestern Polytech Univ Sch Automat Xian 710129 Peoples R China

The increasing demand for a better quality of service in radar and wireless communications has attracted research on interference suppression. Nonlinear filtering methods based on spline architecture have been widely utilized in signal processing due to their effectiveness against alpha-stable interference. However, existing nonlinear spline adaptive filter algorithms suffer from high steady-state misalignment. To achieve lower steady-state misalignment along with having comparable computational complexity, we propose a nonlinear spline Versoria prioritization optimization adaptive filter (SPOAF-MVC) for alpha-stable clutter in this article. Furthermore, we study the bound on learning rate and computational complexity for the proposed algorithm. Numerical simulations confirm the effectiveness and efficiency of the proposed SPOAF-MVC algorithm for alpha-stable clutter under the Wiener system identification.

关键词： Adaptive filters Clutter Splines (mathematics) Kernel signal processing algorithms Robustness Computational complexity Alpha-stable clutter maximum Versoria criterion (MVC) nonlinear adaptive filter Wiener system identification

来源：评论

学校读者我要写书评

暂无评论

Multivariate Dynamic Mode Decomposition and Its Application to Bearing Fault Diagnosis

引用

IEEE SENSORS JOURNAL 2023年第7期23卷 7514-7524页

作者： Zhang, Qixiang Yuan, Rui Lv, Yong Li, Zhaolun Wu, Hongan Wuhan Univ Sci & Technol WUST Key Lab Met Equipment & Control Technol Minist Educ Wuhan 430081 Peoples R China Wuhan Univ Sci & Technol WUST Hubei Key Lab Mech Transmiss & Mfg Engn Wuhan 430081 Peoples R China

In practical engineering applications, the multivariate signal contains more fault feature information than the single-channel signal. How to realize synchronous extraction of fault features from the multivariate signal is of great significance in fault diagnosis of rotary machinery. Dynamic mode decomposition (DMD) has attracted much attention due to its excellent dynamic feature extraction ability. However, DMD lacks mode aliasing property when dealing with the multivariate signal, which may lead to the loss of critical fault feature information. Cater to this problem, this article proposed a multivariate DMD (MDMD) algorithm that is the multivariate extension of DMD. First, snapshot tensors are defined to convert multivariate signals to tensor format. Then, the MDMD algorithm is proposed by introducing tensor operations into the original DMD algorithm, where a tensor low tubal rank component extraction framework is constructed to enable simultaneous extraction of bearing fault features from the multivariate signal, to enhance the robustness and effectiveness of the algorithm. Finally, both numerical simulations and experiments verify that the proposed MDMD has higher fault diagnosis accuracy than other multivariate signal-processing methods.

关键词： Feature extraction Tensors Heuristic algorithms Fault diagnosis signal processing algorithms Matrix decomposition Sensors Bearing fault diagnosis multivariate dynamic mode decomposition (MDMD) multivariate signal processing transient feature extraction

来源：评论

学校读者我要写书评

暂无评论

Model-Driven Based Deep Unfolding Equalizer for Underwater Acoustic OFDM Communications

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2023年第5期72卷 6056-6067页

作者： Zhao, Hao Yang, Cui Xu, Yalu Ji, Fei Wen, Miaowen Chen, Yankun South China Univ Technol Sch Elect & Informat Engn Guangzhou 510640 Peoples R China South China Univ Technol Sch Elect & Informat Engn Guangzhou 510640 Peoples R China Guangdong Prov key Lab short rangewireless detect Guangzhou 510640 Peoples R China Minist Nat Resources Guangdong Prov key Lab short rangewireless detect Key Lab Marine Environm Survey Technol & Applicat Guangzhou 510640 Peoples R China

It is challenging to design an equalizer for the complex time-frequency doubly-selective channel. In this paper, we employ the deep unfolding approach to establish an equalizer for the underwater acoustic (UWA) orthogonal frequency division multiplexing (OFDM) system, namely UDNet. Each layer of UDNet is designed according to the classical minimum mean square error (MMSE) equalizer. Moreover, we consider the QPSK equalization as a four-classification task and adopt minimum Kullback-Leibler (KL) to achieve a smaller symbol error rate (SER) with the one-hot coding instead of the MMSE criterion. In addition, we introduce a sliding structure based on the banded approximation of the channel matrix to reduce the network size and aid UDNet to perform well for different-length signals without changing the network structure. Furthermore, we apply the measured at-sea doubly-selective UWA channel and offshore background noise to evaluate the proposed equalizer. Experimental results show that the proposed UDNet performs better with low computational complexity. Concretely, the SER of UDNet is nearly an order of magnitude lower than that of MMSE.

关键词： Equalizers OFDM Symbols signal processing algorithms Receivers Heating systems Underwater acoustics Unfolding-based equalizer deep learning doubly-selective underwater acoustic communication

来源：评论

学校读者我要写书评

暂无评论

Plug-and-Play Methods for Integrating Physical and Learned Models in Computational Imaging: Theory, algorithms, and applications

引用

IEEE signal processing MAGAZINE 2023年第1期40卷 85-97页

作者： Kamilov, Ulugbek S. Bouman, Charles A. Buzzard, Gregery T. Wohlberg, Brendt Washington Univ Computat Imaging Grp St Louis MO USA Purdue Univ Elect & Comp Engn & Biomed Engn W Lafayette IN USA Purdue Univ Math Fac W Lafayette IN 47907 USA Los Alamos Natl Lab Theoret Div Los Alamos NM USA

Plug-and-play (PnP) priors constitute one of the most widely used frameworks for solving computational imaging problems through the integration of physical models and learned models. PnP leverages high-fidelity physical sensor models and powerful machine learning methods for prior modeling of data to provide state-of-the-art reconstruction algorithms. PnP algorithms alternate between minimizing a data fidelity term to promote data consistency and imposing a learned regularizer in the form of an image denoiser. Recent highly successful applications of PnP algorithms include biomicroscopy, computerized tomography (CT), magnetic resonance imaging (MRI), and joint ptychotomography. This article presents a unified and principled review of PnP by tracing its roots, describing its major variations, summarizing main results, and discussing applications in computational imaging. We also point the way toward further developments by discussing recent results on equilibrium equations that formulate the problem associated with PnP algorithms.

关键词： Training data Machine learning algorithms Computational modeling Magnetic resonance imaging Computed tomography Imaging signal processing algorithms

来源：评论

学校读者我要写书评

暂无评论

Joint Analog and Digital Self-Interference Cancellation for Full Duplex Transceiver With Frequency-Dependent I/Q Imbalance

引用

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS 2023年第4期22卷 2364-2378页

作者： Huang, Xiaojing Le, Anh Tuyen Guo, Y. Jay Univ Technol Sydney Global Big Data Technol Ctr Ultimo NSW 2007 Australia

An effective and practical joint analog and digital self-interference cancellation (SIC) scheme without additional signalling overhead for an I/Q imbalanced full duplex transceiver is proposed in this paper. This scheme combines an I/Q imbalanced analog least mean square (ALMS) loop at the transceiver radio frequency frontend and a two-stage digital signal processing (DSP) at the digital baseband to achieve excellent SIC performance with low complexity. The steady state weighting coefficients of the I/Q imbalanced ALMS loop with periodical transmitted signal and the loop's convergence behaviour are firstly analysed. The residual SI is then modelled as the output of a time-varying widely linear system. With a track/hold control mechanism applied to the ALMS loop, the system model for digital SIC is further presented, followed by the DSP algorithms suitable for real-time implementation. The noise enhancement in each stage digital cancellation is also analysed and formulated. Finally, simulation results are provided to verify the theoretical analyses and demonstrate the overall SIC performance.

关键词： Interference cancellation Transceivers signal processing algorithms Wireless communication Tracking loops RF signals Mathematical models In-band full duplex (IBFD) self-interference cancellation (SIC) analog least mean square (ALMS) loop joint analog and digital SIC

来源：评论

学校读者我要写书评

暂无评论

Bayesian Deep Learning for Image Reconstruction: From structured sparsity to uncertainty estimation

引用

IEEE signal processing MAGAZINE 2023年第1期40卷 73-84页

作者： Dong, Weisheng Wu, Jinjian Li, Leida Shi, Guangming Li, Xin Hong Kong Polytech Univ Dept Comp Hong Kong Peoples R China Xidian Univ Sch Artificial Intelligence Xian 710071 Peoples R China Nanyang Technol Univ Singapore Singapore Xidian Univ Sch Artificial Intelligence Xian 710071 Peoples R China Nanyang Technol Univ Sch Elect & Elect Engn Rapid Rich Object Search Lab Singapore Singapore Xidian Univ Sch Elect Engn Xian Peoples R China Xidian Univ Sch Artificial Intelligence Xian 710071 Peoples R China Univ Sci & Technol China Elect Engn & informat Sci Hefei Peoples R China West Virginia Univ Lane Dept Comp Sci & Elect Engn Morgantown WV 26506 USA

Conventional wisdom in model-based computational imaging incorporates physics-based imaging models, noise characteristics, and image priors into a unified Bayesian framework. Rapid advances in deep learning have inspired a new generation of data-driven computational imaging systems with performances even better than those of their model-based counterparts. However, the design of learning-based algorithms for computational imaging often lacks transparency, making it difficult to optimize the entire imaging system in a complete manner.

关键词： Deep learning Uncertainty Image coding Computational modeling Imaging signal processing algorithms Network architecture

来源：评论

学校读者我要写书评

暂无评论

State-of-the-Art Analysis of Deep Learning-Based Monaural Speech Source Separation Techniques

引用

IEEE ACCESS 2023年 11卷 4242-4269页

作者： Soni, Swati Yadav, Ram Narayan Gupta, Lalita Maulana Azad Natl Inst Technol Dept Elect & Commun Engn Bhopal 462003 Madhya Pradesh India

The monaural speech source separation problem is an important application in the signal pro-cessing field. But recent interaction of deep learning algorithms with signal processing achieves remarkable performance improvement for speech source separation problems. This paper explores the numerous state-of-the-art deep learning-based monaural speech source separation algorithms in the time-frequency (T-F), time, and hybrid domains. The motivation, algorithm, and framework of different deep learning models for monaural speech source separation are analyzed. The benchmarked algorithms in the T-F domain can be categorized as deep neural networks (DNN), clustering, permutation, multi-task learning, computational auditory sense analysis (CASA), and phase reconstruction-based techniques, whereas the state-of-the-art time-domain approaches can be categorized as CNN, RNN, multi-scale fusion (MSF), and transformer-based techniques. The end-to-end post filter (E2EPF) is a hybrid algorithm combining T-F and time-domain works to achieve enhanced results. Time-domain models have shown improvement in separation performance compared to the T-F and hybrid domain models with small model sizes. Methods in T-F, time, and hybrid domains are compared using SDR, SI - SDR, SI - SNR, PESQ, and STOI as quality assessment metrics on some benchmark datasets.

关键词： Source separation Hidden Markov models Time-domain analysis Speech enhancement Deep learning signal processing algorithms Computational modeling Deep-clustering deep learning monaural speech source separation permutation invariant training time domain speaker separation

来源：评论

学校读者我要写书评

暂无评论

Altimetric Parameter Estimation in Long Coherent processing by Airborne Delay/Doppler Altimetry Based on Bayesian Learning

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2023年 16卷 8135-8148页

作者： Fang, Cheng Sun, Ming Huang, Bo Guan, Fangheng Zhou, Honghao Liao, Xianhua Yang, Lei Civil Aviat Univ China Tianjin Key Lab Adv Signal Proc Tianjin 300300 Peoples R China China Acad Engn Phys Inst Elect Engn Mianyang 621900 Peoples R China

The along-track resolution of conventional airborne synthetic aperture radar altimetry (SARAL), where the coherent processing interval (CPI) is the burst length, is not a full resolution due to beam limited. For a fully focused radar altimeter image, a novel airborne SARAL processing algorithm based on a long CPI is proposed in this article. The long CPI is capable of achieving higher azimuthal resolution than that in the conventional altimeter with limited number of pulses. To conquer the problem of atmosphere turbulence and in accordance of motion deviation of the airborne platform, a subaperture phase gradient autofocusing framework is introduced to alleviate nonsystematic phase errors (NsPE). In this framework, NsPE is estimated and compensated in along-track so that a fully focused delayed Doppler map (DDM) can be guaranteed. Finally, the height parameter estimation is performed to the 1-D altimeter echoes after multilooking processing of DDM to improve the estimation accuracy. Given that the conventional retracking algorithm is sensitive to noise, which may degrade the estimation accuracy, a flexible Bayesian method is designed in a hierarchical manner for the SARAL retracking. The SARAL raw data are utilized in the experiment. The results of image entropy and RMSE demonstrate the effectiveness and superiority of the proposed algorithm both qualitatively and quantitatively.

关键词： Delays Electronics packaging Doppler effect Altimetry signal processing algorithms Bayes methods Synthetic aperture radar Coherent processing interval (CPI) hierarchical Bayesian learning phase gradient autofocusing (PGA) synthetic aperture radar altimeter (SARAL)

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：