检索结果-内蒙古大学图书馆

25th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)

作者： Hu, Rui Liu, Huafeng Zhejiang Univ Dept Opt Engn State Key Lab Modern Opt Instrumentat Hangzhou 310027 Peoples R China Jiaxing Key Lab Photon Sensing Intelligent Imagin Jiaxing 314000 Peoples R China Zhejiang Univ Jiaxing Res Inst Intelligent Opt & Photon Res Ctr Jiaxing 314000 Peoples R China

ISBN: (纸本)9783031164408;9783031164392

Positron emission tomography (PET) image reconstruction is an ill-posed inverse problem and suffers from high level of noise due to limited counts received. Recently deep neural networks especially convolutional neural networks (CNN) have been successfully applied to PET image reconstruction. However, the local characteristics of the convolution operator potentially limit the image quality obtained by current CNN-based PET image reconstruction methods. In this paper, we propose a residual swin-transformer based regularizer (RSTR) to incorporate regularization into the iterative reconstruction framework. Specifically, a convolution layer is firstly adopted to extract shallow features, then the deep feature extraction is accomplished by the swin-transformer layer. At last, both deep and shallow features are fused with a residual operation and another convolution layer. Validations on the realistic 3D brain simulated low-count data show that our proposed method outperforms the state-of-the-art methods in both qualitative and quantitative measures.

关键词： Positron Emission Tomography (PET) Image reconstruction model-based deep learning Transformer

来源：评论

学校读者我要写书评

暂无评论

MONOTONICALLY CONVERGENT REGULARIZATION BY DENOISING 29

MONOTONICALLY CONVERGENT REGULARIZATION BY DENOISING

引用

IEEE International Conference on Image Processing (ICIP)

作者： Hu, Yuyang Liu, Jiaming Xu, Xiaojian Kamilov, Ulugbek S. Washington Univ St Louis Dept Elect & Syst Engn St Louis MO 63130 USA Washington Univ St Louis Dept Comp Sci & Engn St Louis MO 63130 USA

ISBN: (数字)9781665496209

ISBN: (纸本)9781665496209

Regularization by denoising (RED) is a widely-used framework for solving inverse problems by leveraging image denoisers as image priors. Recent work has reported the stateof-the-art performance of RED in a number of imaging applications using pre-trained deep neural nets as denoisers. Despite the recent progress, the stable convergence of RED algorithms remains an open problem. The existing RED theory only guarantees stability for convex data-fidelity terms and nonexpansive denoisers. This work addresses this issue by developing a new monotone RED (MRED) algorithm, whose convergence does not require nonexpansiveness of the deep denoising prior. Simulations on image deblurring and compressive sensing recovery from random matrices show the stability of MRED even when the traditional RED diverges.

关键词： Inverse problems plug-and-play priors regularization by denoising model-based deep learning

来源：评论

学校读者我要写书评

暂无评论

SPATIAL-SPECTRAL CONVOLUTIONAL SPARSE NEURAL NETWORK FOR HYPERSPECTRAL IMAGE DENOISING

SPATIAL-SPECTRAL CONVOLUTIONAL SPARSE NEURAL NETWORK FOR HYP...

引用

IEEE International Geoscience and Remote Sensing Symposium (IGARSS)

作者： Xiong, Fengchao Ye, Minchao Zhou, Jun Qian, Yuntao Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing Peoples R China China Jiliang Univ Coll Informat Engn Hangzhou Peoples R China Griffith Univ Sch Informat & Commun Technol Nathan Qld Australia Zhejiang Univ Coll Comp Sci Hangzhou Peoples R China

ISBN: (数字)9781665427920

ISBN: (纸本)9781665427920

Sparse representation (SR) is a widely accepted hyperspectral image (HSI) denoising model. Because of the curse of dimensionality and the desire to better fit the data, the SR models are typically deployed on small and fully overlapping blocks whose results are averaged to produce the global denoised HSI. This "local-global" denoising mechanism ignores the dependencies between blocks, resulting in visual artifacts. This paper describes the underlying clean HSI with a 3D convolutional sparse coding (CSC) model, representing the HSI with a linear combination of few shift-invariant 3D spatial-spectral filters in a global dictionary. Instead of operating on patches, the CSC model sees the clean HSI is generated from a sum of local atoms that appear in a small number of locations throughout the image, naturally retaining the relationship between pixels. Moreover, we unfold the optimization process of the model into a spatial-spectral convolutional sparse neural network which absorbs the interpretation ability of the model while supporting discriminative learning from data. Experimental results on both synthetic and real-world datasets show that our network achieves competitive denoising performances, qualitatively and quantitatively.

关键词： Hyperspectral image denoising model-based deep learning convolutional sparse coding deep unfolding

来源：评论

学校读者我要写书评

暂无评论

LoRD-Net: Unfolded deep Detection Network With Low-Resolution Receivers

引用

IEEE TRANSACTIONS ON SIGNAL PROCESSING 2021年 69卷 5651-5664页

作者： Khobahi, Shahin Shlezinger, Nir Soltanalian, Mojtaba Eldar, Yonina C. Univ Illinois ECE Dept Chicago IL 60607 USA Ben Gurion Univ Negev Sch ECE IL-84105 Beer Sheva Israel Weizmann Inst Sci Fac Math & CS IL-7610001 Rehovot Israel

The need to recover high-dimensional signals from their noisy low-resolution quantized measurements is widely encountered in communications and sensing. In this paper, we focus on the extreme case of one-bit quantizers, and propose a deep detector entitled LoRD-Net for recovering information symbols from one-bit measurements. Our method is a model-aware data-driven architecture based on deep unfolding of first-order optimization iterations. LoRD-Net has a task-based architecture dedicated to recovering the underlying signal of interest from the one-bit noisy measurements without requiring prior knowledge of the channel matrix through which the one-bit measurements are obtained. The proposed deep detector has much fewer parameters compared to black-box deep networks due to the incorporation of domain-knowledge in the design of its architecture, allowing it to operate in a data-driven fashion while benefiting from the flexibility, versatility, and reliability of model-based optimization methods. LoRD-Net operates in a blind fashion, which requires addressing both the non-linear nature of the data-acquisition system as well as identifying a proper optimization objective for signal recovery. Accordingly, we propose a two-stage training method for LoRD-Net, in which the first stage is dedicated to identifying the proper form of the optimization process to unfold, while the latter trains the resulting model in an end-to-end manner. We numerically evaluate the proposed receiver architecture for one-bit signal recovery in wireless communications and demonstrate that the proposed hybrid methodology outperforms both data-driven and model-based state-of-the-art methods, while utilizing small datasets, on the order of merely similar to 500 samples, for training.

关键词： Receivers Task analysis Computational modeling Training Quantization (signal) Maximum likelihood estimation Detectors model-based deep learning deep unfolding machine learning massive MIMO low-resolution signal processing one-bit quantization

来源：评论

学校读者我要写书评

暂无评论

Graph Convolutional Networks for model-based learning in Nonlinear Inverse Problems

引用

IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING 2021年 7卷 1341-1353页

作者： Herzberg, William Rowe, Daniel B. Hauptmann, Andreas Hamilton, Sarah J. Marquette Univ Dept Math & Stat Sci Milwaukee WI 53233 USA Univ Oulu Res Unit Math Sci Oulu 90570 Finland UCL Dept Comp Sci London WC1E 6BT England

The majority of model-based learned image reconstruction methods in medical imaging have been limited to uniform domains, such as pixelated images. If the underlying model is solved on nonuniform meshes, arising from a finite element method typical for nonlinear inverse problems, interpolation and embeddings are needed. To overcome this, we present a flexible framework to extend model-based learning directly to nonuniform meshes, by interpreting the mesh as a graph and formulating our network architectures using graph convolutional neural networks. This gives rise to the proposed iterative Graph Convolutional Newton-type Method (GCNM), which includes the forward model in the solution of the inverse problem, while all updates are directly computed by the network on the problem specific mesh. We present results for Electrical Impedance Tomography, a severely ill-posed nonlinear inverse problem that is frequently solved via optimization-based methods, where the forward problem is solved by finite element methods. Results for absolute EIT imaging are compared to standard iterative methods as well as a graph residual network. We show that the GCNM has good generalizability to different domain shapes andmeshes, out of distribution data as well as experimental data, from purely simulated training data and without transfer training.

关键词： Finite element method graph convolutional networks model-based deep learning conductivity electrical impedance tomography

来源：评论

学校读者我要写书评

暂无评论

Dynamic low-count PET image reconstruction using spatio-temporal primal dual network

引用

PHYSICS IN MEDICINE AND BIOLOGY 2023年第13期68卷 135015-135015页

作者： Hu, Rui Cui, Jianan Li, Chenxu Yu, Chengjin Chen, Yunmei Liu, Huafeng Zhejiang Univ Dept Opt Engn State Key Lab Modern Opt Instrumentat Hangzhou 310027 Peoples R China Zhejiang Univ Technol Inst Informat Proc & Automat Coll Informat Engn Hangzhou 310001 Peoples R China Univ Florida Dept Math Gainesville FL 32611 USA

Objective. Dynamic positron emission tomography (PET) imaging, which can provide information on dynamic changes in physiological metabolism, is now widely used in clinical diagnosis and cancer treatment. However, the reconstruction from dynamic data is extremely challenging due to the limited counts received in individual frame, especially in ultra short frames. Recently, the unrolled model-based deep learning methods have shown inspiring results for low-count PET image reconstruction with good interpretability. Nevertheless, the existing model-based deep learning methods mainly focus on the spatial correlations while ignore the temporal domain. Approach. In this paper, inspired by the learned primal dual (LPD) algorithm, we propose the spatio-temporal primal dual network (STPDnet) for dynamic low-count PET image reconstruction. Both spatial and temporal correlations are encoded by 3D convolution operators. The physical projection of PET is embedded in the iterative learning process of the network, which provides the physical constraints and enhances interpretability. Main results. The experiments of both simulation data and real rat scan data have shown that the proposed method can achieve substantial noise reduction in both temporal and spatial domains and outperform the maximum likelihood expectation maximization, spatio-temporal kernel method, LPD and FBPnet. Significance. Experimental results show STPDnet better reconstruction performance in the low count situation, which makes the proposed method particularly suitable in whole-body dynamic imaging and parametric PET imaging that require extreme short frames and usually suffer from high level of noise.

关键词： image reconstruction positron emission tomography model-based deep learning low count spatio-temporal correlation

来源：评论

学校读者我要写书评

暂无评论

Attentional filtering: Dynamic system state estimation as an attention

引用

INFORMATION FUSION 2025年 123卷

作者： Lin, Andi Zhang, Wen-An Guo, Lei Zhejiang Univ Technol Coll Informat Engn Hangzhou 310023 Peoples R China Beihang Univ Sch Automat Sci & Elect Engn Beijing 100191 Peoples R China

The ability to understand the behavior of dynamic systems is supported by state estimation techniques. Traditional methods for estimating states involve developing a probabilistic paradigm using a generic state-space representation, propagating distributions within the Bayesian framework. Due to highly nonlinear dynamics and rich sensory measurements, a fundamental bottleneck is argued to naturally emerge within Bayesian probabilistic estimators. The fundamental bottleneck stems from the intractability of exact posterior inference in complex settings. The distributions fusion for estimation purposes requires Bayesian inference based on the assumption of a specific distribution (typically Gaussian), where Bayesian inference generates only a poor approximation. To go beyond Gaussian filtering, a novel filtering framework named "Attentional Filtering" (AF) for the state estimation is framed. This framework, guided by the human attention mechanism, considers state estimation as two behavioral phenomena of sensory attention. Additionally, a specific computational model with deep learning is established for AF framework and applied to both simulation data and real-world datasets. Extensive experiments validate excellent performance and robustness of the proposed AF in complex settings. The proposed framework provides the ability to estimate states of arbitrary dynamic systems, giving potentially novel insights at state estimation.

关键词： State estimation model-based deep learning Attention mechanism State-space model Bayesian filter

来源：评论

学校读者我要写书评

暂无评论

Dual-channel prior-based deep unfolding with contrastive learning for underwater image enhancement

引用

Journal of Visual Communication and Image Representation 2025年 111卷

作者： Thuy Thi Pham Truong Thanh Nhat Mai Hansung Yu Chul Lee Department of Multimedia Engineering Dongguk University Seoul 04620 Republic of Korea MoMo Company Ho Chi Minh City 700000 Vietnam FionSystems Inc. Seongnam 13449 Republic of Korea Department of Computer Science and Artificial Intelligence Dongguk University Seoul 04620 Republic of Korea

Underwater image enhancement (UIE) techniques aim to improve the visual quality of underwater images degraded by wavelength-dependent light absorption and scattering. In this work, we propose a deep unfolding approach for UIE to leverage the advantages of both model- and learning-based approaches while overcoming their weaknesses. Specifically, we first formulate the UIE task as a joint optimization problem with physics-based priors, providing a robust theoretical foundation on the properties of underwater imaging. Then, we define implicit regularizers to compensate for modeling inaccuracies in the physics-based priors and solve the optimization using an iterative technique. Finally, we unfold the iterative algorithm into a series of interconnected blocks, where each block represents a single iteration of the algorithm. We further improve performance by employing a contrastive learning strategy that learns discriminative representations between the underwater and clean images. Experimental results demonstrate that the proposed algorithm provides better enhancement performance than state-of-the-art algorithms.

关键词： Contrastive learning deep unfolding model-based deep learning Underwater image enhancement

来源：评论

学校读者我要写书评

暂无评论

learning SAMPLING AND model-based SIGNAL RECOVERY FOR COMPRESSED SENSING MRI

LEARNING SAMPLING AND MODEL-BASED SIGNAL RECOVERY FOR COMPRE...

引用

IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Huijben, Iris A. M. Veeling, Bastiaan S. van Sloun, Ruud J. G. Eindhoven Univ Technol Dept Elect Engn Eindhoven Netherlands Univ Amsterdam Dept Comp Sci Amsterdam Netherlands

ISBN: (纸本)9781509066315

Compressed sensing (CS) MRI relies on adequate under-sampling of the k-space to accelerate the acquisition without compromising image quality. Consequently, the design of optimal sampling patterns for these k-space coefficients has received significant attention, with many CS MRI methods exploiting variable-density probability distributions. Realizing that an optimal sampling pattern may depend on the downstream task (e.g. image reconstruction, segmentation, or classification), we here propose joint learning of both task-adaptive k-space sampling and a subsequent model-based proximal-gradient recovery network. The former is enabled through a probabilistic generative model that leverages the Gumbel-softmax relaxation to sample across trainable beliefs while maintaining differentiability. The proposed combination of a highly flexible sampling model and a model-based (sampling-adaptive) image reconstruction network facilitates exploration and efficient training, yielding improved MR image quality compared to other sampling baselines.

关键词： Compressed sensing model-based deep learning magnetic resonance imaging

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：