检索结果-内蒙古大学图书馆

A Physics-Assisted Deep-Learning Scheme Based on Globally Perceptive Modules for Electromagnetic Inverse Scattering Problems

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2025年 63卷

作者： Du, Changlin Pan, Jin Yang, Deqiang Hu, Jun Nie, Zaiping Chen, Yongpin Univ Elect Sci & Technol China Sch Elect Sci & Engn Chengdu 611731 Peoples R China

The multiple scattering effect is commonly present in electromagnetic (EM) scattering and can serve as a priori information to guide the solution of inverse scattering problems (ISPs). As depicted by the Green's function, the multiple scattering effect exhibits both global and local features, i.e., the interaction between subscatterers is in principle global but the coupling strength is usually large in the neighboring region and decays with distance. For solving ISPs, the convolutional layer (CL) has been proven to be highly effective as a fundamental building block in learning-based solvers. Through supervised learning, CL-based networks can predict the refined result from the pre-solution of EM parameters. However, due to the limitation of local perception capability, CL can only model the local feature of the multiple scattering effect. To remedy this, this article proposes a globally perceptive (GP) module that cascades a global self-attention layer (GSAL) and a CL to capture both global interaction and local emphasis. To show the effectiveness of the GP module, a deep-learning scheme with global perception (DLSGP) is constructed using a U-shape structure, where the pre-solution is provided by the backpropagation (BP) or domain current (DC) method. Validation results using cross-set data, noisy data, and measured data demonstrate the high accuracy, strong generalization ability, and robustness of the proposed module. The study on the global perception ability further confirms that the proposed module contains more information on the multiple scattering effect. Since this effect is common in ISPs, this fundamental module can be applied to other learning-based solvers.

关键词： Scattering Neural networks Learning systems Accuracy Vectors Iterative methods Inverse problems Electromagnetic scattering Robustness reconstruction algorithms Deep learning global self-attention globally perceptive (GP) module inverse scattering problems (ISPs) multiple scattering effect

来源：评论

学校读者我要写书评

暂无评论

Hybrid Domain Enhancement Network for Ultraclean Pure Shift Spectroscopy With Fast Acquisition

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2025年 74卷

作者： Shen, Jia Huang, Mingkai Chen, Yida Wu, Liubin Li, Hong Yang, Yu Chen, Zhong Xiamen Univ Dept Elect Sci Xiamen 361005 Peoples R China Univ Hong Kong Dept Elect & Elect Engn Hong Kong Peoples R China Jimei Univ Sch Ocean Informat Engn Xiamen 361021 Peoples R China

Pure shift technique can greatly improve the resolution of proton spectra compared to conventional nuclear magnetic resonance (NMR) spectroscopy but at the cost of longer acquisition time. To accelerate data acquisition in pure shift experiments, nonuniform sampling (NUS) is usually incorporated along with advanced data processing methods to reconstruct the spectrum. While some deep learning (DL) approaches have been proposed for this purpose, their performance declines when dealing with data under varying experimental conditions, primarily because they solely rely on frequency-domain information. To overcome this limitation, we develop a novel hybrid domain enhancement (HDE) network that combines time- and frequency-domain signals as input. We introduced an adaptive-domain transform unit (ADTU) to extract time-domain features that capture segmental correlations while retaining some temporal sequential information. Our experiments demonstrate that the proposed network effectively suppresses chunking sidebands and undersampling artifacts, showing exceptional robustness in complex scenarios, particularly under phase distortion. On real experimental data, it achieves a 5%-33% higher precision than state-of-the-art (SOTA) methods, indicating fewer artifacts in the reconstructed spectrum. This approach holds great potential for broader application in NMR spectroscopy, offering clearer and more accurate spectral interpretation.

关键词： Nuclear magnetic resonance Time-domain analysis Image reconstruction Couplings Amplitude modulation Spectroscopy Phase distortion Discrete Fourier transforms Time-frequency analysis reconstruction algorithms Chunking sideband neural network nonuniform sampling (NUS) pure shift signal reconstruction

来源：评论

学校读者我要写书评

暂无评论

Multi-Scale Channel Distillation Network for Image Compressive Sensing

引用

IEEE ACCESS 2025年 13卷 9524-9537页

作者： Zhang, Tianyu Ye, Kuntao Zhang, Yue Lu, Rui Jiangxi Univ Sci & Technol Sch Sci Ganzhou 341000 Peoples R China State Grid Xinyu Power Supply Co Elect Power Commun Dept Xinyu 338000 Peoples R China

Recently, convolutional neural networks (CNNs) have demonstrated striking success in computer vision tasks. Methods based on CNNs for image compressive sensing (CS) have also gained prominence. However, existing methods tend to increase the depth of the network in feature space for better reconstruction quality, neglecting the hierarchical representation of intermediate features in pixel space. In order to coordinate the feature space and pixel space to complete the deep reconstruction of images, and further improve the reconstruction performance of current CS methods, we propose a multi-scale channel distillation network (MSCDN). This network first obtains images of multiple scales using a scale-space image decomposition method at the sampling stage, followed by sampling these decomposed images through a convolutional operation. In this way, multi-scale information in the compressed domain is aggregated. During the reconstruction phase, a low-frequency information recovery network generates a preliminary image, whereas a high-frequency feature aggregation network refines the image further. Specifically, we design a dual-branch deep reconstruction architecture with channel distillation residual block (CDRB) as the core component. One branch extracts features gradually by cascading multiple CDRB modules, thereby supplementing the initial reconstructed image with a large amount of high-frequency content in feature space. The other branch takes the initial reconstructed image as input and sequentially fuses the intermediate feature outputs by CDRBs to increase the local details of the image in pixel space. Combining outputs from both branches, we achieve an optimal reconstructed image. Extensive experimental results on four benchmark datasets demonstrate that MSCDN surpasses state-of-the-art CS methods not only in reconstruction accuracy but also in perceptual visual quality.

关键词： Image reconstruction Feature extraction Convolution reconstruction algorithms Image coding Computational complexity Compressed sensing Training Optimization Matched filters Convolutional neural networks compressive sensing intermediate feature multi-scale image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Acoustic Resolution Photoacoustic Microscopy Imaging Enhancement: Integration of Group Sparsity With Deep Denoiser Prior

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2025年 34卷 522-537页

作者： Zhang, Zhengyuan Pan, Zuozhou Lin, Zhuoyi Sharma, Arunima Lin, Chia-Wen Pramanik, Manojit Zheng, Yuanjin Nanyang Technol Univ Sch Elect & Elect Engn Jurong West 639798 Singapore China Jiliang Univ Coll Metrol & Measurement Engn Hangzhou 314423 Peoples R China ASTAR Inst Infocomm Res Singapore 138632 Singapore Johns Hopkins Univ Dept Elect & Comp Engn Baltimore MD 21218 USA Natl Tsing Hua Univ Dept Elect Engn Hsinchu 300 Taiwan Iowa State Univ Dept Elect & Comp Engn Ames IA 50011 USA

Acoustic resolution photoacoustic microscopy (AR-PAM) is a novel medical imaging modality, which can be used for both structural and functional imaging in deep bio-tissue. However, the imaging resolution is degraded and structural details are lost since its dependency on acoustic focusing, which significantly constrains its scope of applications in medical and clinical scenarios. To address the above issue, model-based approaches incorporating traditional analytical prior terms have been employed, making it challenging to capture finer details of anatomical bio-structures. In this paper, we proposed an innovative prior named group sparsity prior for simultaneous reconstruction, which utilizes the non-local structural similarity between patches extracted from internal AR-PAM images. The local image details and resolution are improved while artifacts are also introduced. To mitigate the artifacts introduced by patch-based reconstruction methods, we further integrate an external image dataset as an extra information provider and consolidate the group sparsity prior with a deep denoiser prior. In this way, complementary information can be exploited to improve reconstruction results. Extensive experiments are conducted to enhance the simulated and in vivo AR-PAM imaging results. Specifically, in the simulated images, the mean peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) values have increased from 16.36 dB and 0.46 to 27.62 dB and 0.92, respectively. The in vivo reconstructed results also demonstrate the proposed method achieves superior local and global perceptual qualities, the metrics of signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) have significantly increased from 10.59 and 8.61 to 30.83 and 27.54, respectively. Additionally, reconstruction fidelity is validated with the optical resolution photoacoustic microscopy (OR-PAM) data as reference image.

关键词： Imaging Image reconstruction Acoustics Biomedical imaging Optical imaging Mathematical models Focusing Biomedical optical imaging Microscopy reconstruction algorithms Acoustic resolution photoacoustic microscopy deep denoiser prior group sparsity prior non-local self-similarity photoacoustic imaging

来源：评论

学校读者我要写书评

暂无评论

Orientation Determination of Cryo-EM Projection Images Using Reliable Common Lines and Spherical Embeddings

引用

IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS 2024年第6期21卷 2496-2509页

作者： Wang, Xiangwen Jin, Qiaoying Zou, Li Lin, Xianghong Lu, Yonggang Northwest Normal Univ Coll Comp Sci & Engn Lanzhou 730070 Peoples R China Lanzhou Univ Hosp 2 Cuiying Biomed Res Ctr Lanzhou 730000 Peoples R China Lanzhou Univ Clin Med Sch Lanzhou 730000 Peoples R China Lanzhou Univ Sch Informat Sci & Engn Lanzhou 730000 Peoples R China Gansu Prov Meteorol Informat & Tech Support & Equi Gansu Meteorol Bur Lanzhou 730020 Peoples R China

Three-dimensional (3D) reconstruction in single-particle cryo-electron microscopy (cryo-EM) is a critical technique for recovering and studying the fine 3D structure of proteins and other biological macromolecules, where the primary issue is to determine the orientations of projection images with high levels of noise. This paper proposes a method to determine the orientations of cryo-EM projection images using reliable common lines and spherical embeddings. First, the reliability of common lines between projection images is evaluated using a weighted voting algorithm based on an iterative improvement technique and binarized weighting. Then, the reliable common lines are used to calculate the normal vectors and local X-axis vectors of projection images after two spherical embeddings. Finally, the orientations of projection images are determined by aligning the results of the two spherical embeddings using an orthogonal constraint. Experimental results on both synthetic and real cryo-EM projection image datasets demonstrate that the proposed method can achieve higher accuracy in estimating the orientations of projection images and higher resolution in reconstructing preliminary 3D structures than some common line-based methods, indicating that the proposed method is effective in single-particle cryo-EM 3D reconstruction.

关键词： Three-dimensional displays Reliability Molecular biophysics Image reconstruction Biology Periodic structures Image resolution Electrons Vectors reconstruction algorithms Angular reconstruction cryo-electron microscopy (cryo-EM) reliable common lines single-particle reconstruction spherical embedding

来源：评论

学校读者我要写书评

暂无评论

Deep-Learning-Based Automated Neuron reconstruction From 3D Microscopy Images Using Synthetic Training Images

引用

IEEE TRANSACTIONS ON MEDICAL IMAGING 2022年第5期41卷 1031-1042页

作者： Chen, Weixun Liu, Min Du, Hao Radojevic, Miroslav Wang, Yaonan Meijering, Erik Hunan Univ Coll Elect & Informat Engn Changsha 410082 Hunan Peoples R China Hunan Univ Natl Engn Lab Robot Visual Percept & Control Tech Changsha 410082 Hunan Peoples R China Nuctech Netherlands Res & Dev Dept NL-3072 AP Rotterdam Netherlands Univ New South Wales Sch Comp Sci & Engn Sydney NSW 2052 Australia

Digital reconstruction of neuronal structures from 3D microscopy images is critical for the quantitative investigation of brain circuits and functions. It is a challenging task that would greatly benefit from automatic neuron reconstruction methods. In this paper, we propose a novel method called SPE-DNR that combines spherical-patches extraction (SPE) and deep-learning for neuron reconstruction (DNR). Based on 2D Convolutional Neural Networks (CNNs) and the intensity distribution features extracted by SPE, it determines the tracing directions and classifies voxels into foreground or background. This way, starting from a set of seed points, it automatically traces the neurite centerlines and determines when to stop tracing. To avoid errors caused by imperfect manual reconstructions, we develop an image synthesizing scheme to generate synthetic training images with exact reconstructions. This scheme simulates 3D microscopy imaging conditions as well as structural defects, such as gaps and abrupt radii changes, to improve the visual realism of the synthetic images. To demonstrate the applicability and generalizability of SPE-DNR, we test it on 67 real 3D neuron microscopy images from three datasets. The experimental results show that the proposed SPE-DNR method is robust and competitive compared with other state-of-the-art neuron reconstruction methods.

关键词： Neurons Image reconstruction Three-dimensional displays Microscopy Training reconstruction algorithms Feature extraction 3D neuron reconstruction neuron morphology deep learning microscopy images

来源：评论

学校读者我要写书评

暂无评论

Compressed Spectrum Sensing of Sparse Wideband Signals Based on Deep Learning

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2024年第6期73卷 8434-8444页

作者： Mei, Ruru Wang, Zhugang Chinese Acad Sci Natl Space Sci Ctr Key Lab Elect & Informat Technol Space Syst Beijing 100190 Peoples R China Univ Chinese Acad Sci Sch Electromagnet Field & Microwave Technol Beijing 100049 Peoples R China Chinese Acad Sci Natl Space Sci Ctr Beijing 100190 Peoples R China

In cognitive radio (CR), compressed spectrum sensing (CSS) is an attractive technology for detecting spectrum holes over a large band with a sub-Nyquist sampling rate. Traditional compressed spectrum (CS) reconstruction methods exhibit poor reconstruction performance and high complexity, due to the unknown and dynamically changing sparsity order of the observed signal, the hand-crafted optimization parameters involved in solvers, and dozens or hundreds of iterations. To cope with these challenges, we propose a wideband signal CS reconstruction algorithm by merging the iterative shrinkage thresholding algorithm (ISTA) and deep learning (DL) through deep unfolding and refer to it as the ISTA deep reconstruction network (ISTADR-Net). The parameters in ISTADR-Net (e.g., regularization and step size parameters) are learned end-to-end rather than being adjusted manually. Next, we propose a convolutional neural network (CNN)-based wideband spectrum sensing (WSS) network (CNNWSS-Net), which makes full use of the sparse characteristics of the reconstructed signal in the frequency domain to improve the WSS performance. The simulation results show that the proposed CSS scheme combining ISTADR-Net and CNNWSS-Net works well in low and high compression ratio scenarios, low and high signal-to-noise ratio (SNR) regions. Compared to the traditional CSS methods and other DL-based CSS methods, the proposed CSS scheme yields superior performance in terms of the probability of detection and the probability of false alarm. Furthermore, the computational complexity and computation time of the proposed CSS scheme are reduced compared to the traditional CSS methods.

关键词： Image reconstruction Sensors Wideband reconstruction algorithms Matching pursuit algorithms Iterative algorithms Frequency-domain analysis Cognitive radio compressed sensing convolutional neural network deep unfolding wideband spectrum sensing

来源：评论

学校读者我要写书评

暂无评论

Temporal Information-Guided Generative Adversarial Networks for Stimuli Image reconstruction From Human Brain Activities

引用

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS 2022年第3期14卷 1104-1118页

作者： Huang, Shuo Sun, Liang Yousefnezhad, Muhammad Wang, Meiling Zhang, Daoqiang Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol MIIT Key Lab Pattern Anal & Machine Intelligence Nanjing 211106 Peoples R China

Understanding how the human brain works has attracted increasing attention in both fields of neuroscience and machine learning. Previous studies use autoencoder and generative adversarial networks (GANs) to improve the quality of stimuli image reconstruction from functional magnetic resonance imaging (fMRI) data. However, these methods mainly focus on acquiring relevant features between two different modalities of data, i.e., stimuli images and fMRI, while ignoring the temporal information of fMRI data, thus leading to suboptimal performance. To address this issue, in this article, we propose a temporal information-guided GAN (TIGAN) to reconstruct visual stimuli from human brain activities. Specifically, the proposed method consists of three key components, including: 1) an image encoder for mapping the stimuli images into latent space;2) a long short-term memory (LSTM) generator for fMRI feature mapping, which is used to capture temporal information in fMRI data;and 3) a discriminator for image reconstruction, which is used to make the reconstructed image more similar to the original image. In addition, to better measure the relationship of two different modalities of data (i.e., fMRI and natural images), we leverage a pairwise ranking loss to rank the stimuli images and fMRI to ensure strongly associated pairs at the top and weakly related ones at the bottom. The experimental results on real-world data sets suggest that the proposed TIGAN achieves better performance in comparison with several state-of-the-art image reconstruction approaches.

关键词： Visualization Neuroscience Functional magnetic resonance imaging reconstruction algorithms Generative adversarial networks Brain modeling Loss measurement Functional magnetic resonance imaging (fMRI) generative adversarial networks (GANs) long-short term memory stimuli image reconstruction

来源：评论

学校读者我要写书评

暂无评论

FPGA Implementation of Staged Projection Refining Multiple Orthogonal Matching Pursuit Algorithm for Compressed Sensing

引用

IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS 2025年第5期33卷 1334-1347页

作者： Liu, Sujuan Liang, Yichen Zhang, Zixing Wan, Peiyuan Beijing Univ Technol Coll Microelect Beijing 100124 Peoples R China

reconstruction algorithms are an integral part of compressed sensing (CS) theory, which can reliably reconstruct the original signal from the low-dimensional compressed signal. The orthogonal matching pursuit (OMP) algorithm has been widely studied and extensively selected in hardware implementations. However, the low reconstruction success rate of the OMP algorithm under high sparsity conditions has led to the proposal and application of more reconstruction algorithms in hardware implementations. In this article, a staged projection refining multiple OMP (SPR-MOMP) algorithm is proposed based on the OMP algorithm. This algorithm improves the reconstruction accuracy by refining the support set using a staged backtracking strategy. It also employs a multiple-atom selection strategy for parallel expansion of the support set, ensuring reconstruction efficiency. The reconstruction simulation demonstrates that the SPR-MOMP algorithm achieves a higher reconstruction success rate than the OMP algorithm, with fewer iterations. A hardware architecture applying the SPR-MOMP algorithm is designed and implemented on a Virtex UltraScale+ field-programmable gate array (FPGA) with N = 1024, M = 256, and K = 36. The proposed architecture achieves a reconstruction signal-to-noise ratio (RSNR) of 44.27 dB, with 20-bit data width and 15-bit fractional width. The maximum clock frequency of the architecture is 200 MHz, enabling reconstruction within 276.6 mu s. The proposed architecture achieves a lower dynamic power consumption of 1929 mW.

关键词： Computer architecture Matching pursuit algorithms Hardware Field programmable gate arrays Image reconstruction Heuristic algorithms Backtracking reconstruction algorithms Signal processing algorithms Sensors Compressed sensing (CS) field-programmable gate array (FPGA) hardware reconstruction orthogonal matching pursuit (OMP) QR decomposition (QRD)

来源：评论

学校读者我要写书评

暂无评论

Profile analysis with reconstruction robustness for measurement data subject to outliers

引用

APPLIED OPTICS 2022年第13期61卷 3777-3785页

作者： Gu, Tianqi Xiong, Cui Tang, Dawei Chen, Jianxiong Lin, Shuwen Fuzhou Univ Sch Mech Engn & Automat Fuzhou 350108 Peoples R China Univ Huddersfield Ctr Precis Technol Huddersfield HD1 3DH W Yorkshire England

In the surface profile analysis, there are often a few observations that contain outliers. Due to the existence of outliers, the application of non-robust reconstruction algorithms for measurement data will become a huge problem because these methods are often sensitive to outliers and the approximation effectiveness will be greatly aggravated. In view of this, this paper presents a novel angle-based moving total least squares reconstruction method, to the best of our knowledge, that applies two-step pre-treatment to handle outliers. The first step is an abnormal point detection process that characterizes the geometric features of discrete points in the support domain through a new angle-based parameter constructed by total least square. Then, the point with the largest anomaly degree is removed, and a relevant weight function is defined to adjust the weights of the remaining points. After pretreatment, the final estimates are calculated by weighted total least squares (WTLS) based on the compact weight function. The detection and removal of outliers are automatic, and there is no need to set a threshold value artificially, which effectively avoids the adverse impacts of human operation. Numerical simulations and experiments verify the applicability of the proposed algorithm as well as its accuracy and robustness. (C) 2022 Optica Publishing Group

关键词： Interpolation Numerical simulation reconstruction algorithms Surfaces White light White light interferometry

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：