检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Liu, Xiaohan Pang, Yanwei Sun, Xuebin Liu, Yiming Hou, Yonghong Wang, Zhenchang Li, Xuelong Tianjin Key Laboratory of Brain-Inspired Intelligence Technology School of Electrical and Information Engineering Tianjin University Tianjin300072 China Tiandatz Technology Co. Ltd. Tianjin301723 China Beijing Friendship Hospital Capital Medical University Beijing10050 China School of Computer Science Center for OPTical IMagery Analysis and Learning Northwestern Polytechnical University Shaanxi Xi’an710072 China

Partial scan is a common approach to accelerate Magnetic Resonance Imaging (MRI) data acquisition in both 2D and 3D settings. However, accurately reconstructing images from partial scan data (i.e., incomplete k-space matrices) remains challenging due to lack of an effectively global receptive field in both spatial and k-space domains. To address this problem, we propose the following: (1) a novel convolutional operator called Faster Fourier Convolution (FasterFC) to replace the two consecutive convolution operations typically used in convolutional neural networks (e.g., U-Net, ResNet). Based on the spectral convolution theorem in Fourier theory, FasterFC employs alternating kernels of size 1×1 (1×1×1 in 3D case) in different domains to extend the dual-domain receptive field to the global and achieves faster calculation speed than traditional Fast Fourier Convolution (FFC). (2) A 2D accelerated MRI method, FasterFC-End-to-End-VarNet, which uses FasterFC to improve the sensitivity maps and reconstruction quality. (3) A multi-stage 3D accelerated MRI method called FasterFC-based Single-to-group Network (FAS-Net) that utilizes a single-to-group algorithm to guide k-space domain reconstruction, followed by FasterFC-based cascaded convolutional neural networks to expand the effective receptive field in the dual-domain. Experimental results on the fastMRI and Stanford MRI Data datasets demonstrate that FasterFC improves the quality of both 2D and 3D reconstruction. Moreover, FAS-Net, as a 3D high-resolution (320×320×256) multi-coil (eight) accelerated MRI method, achieves superior reconstruction performance in both qualitative and quantitative results compared with state-of-the-art 2D and 3D methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Investigating the effects of updating rules on cooperation by incorporating interactive diversity

引用

The European Physical Journal B 2021年第2期94卷 1-8页

作者： Zhu, Peican Hou, Xin Guo, Yangming Xu, Jiwei Liu, Jinzhuo School of Computer Science and Engineering Northwestern Polytechnical University Xi’an China Center for OPTical IMagery Analysis and Learning (OPTIMAL) Northwestern Polytechnical University Xi’an China School of Communication and Information Engineering Xi’an University of Posts and Telecommunications Xi’an China School of Software Engineering Yunnan University Kunming China

In reality, most individuals are prone to vary strategies when interacting with their counterparts; thus, it is highly likely that different strategies will be applied when confronting with different players, being referred to as interactive diversity. Numerous scholars have devoted their endless efforts into the investigation of the emergence and maintenance of cooperation for interactive diversity scenarios and this becomes an interesting research topic recently. However, evolutionary dynamics of such games still needs to be further studied. Here, a co-evolving mechanism is proposed aiming to study the effects of applying different updating rules on the level of cooperation when considering interactive diversity. Teaching and learning updating rules are considered in the co-evolving mechanism. Then, we have done extensive experiments and corresponding simulation results are provided. Sufficient analyses of the simulation results are given in order to understand the origin of the observed experimental phenomena. We find the fact that with the increase of the proportion of Type-T players, individuals start to adopt the strategy of cooperation even if the temptation to defect is relatively large; this indicates players are inclined to cooperate under such scenario. All in all, we hope the findings in this manuscript are capable of providing some valuable and interesting insights to solve the social dilemmas.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Meta-Relation Network for Visual Few-Shot learning

Deep Meta-Relation Network for Visual Few-Shot Learning

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Fahong Zhang Qi Wang Xuelong Li School of Computer Science and Center for OPTical IMagery Analysis and Learning Northwestern Polytechnical University Xi’an Shaanxi China

ISBN: (数字)9781509066315

ISBN: (纸本)9781509066322

This paper proposes a novel metric-based deep learning method to solve the few-shot learning problem. It models the relation between images as high dimensional vector, and trains a network module to judge, when given two relational features, which one indicates a stronger connection between the image objects. By training such a network module, we introduce a comparative mechanism into the metric space, i.e., the similarity score of any two images is computed after seeing other images in the same task. Further more, we propose to incorporate a batch classification loss into episodic training to mitigate the hard training problem that occurs when embedding network is going deeper. Experiments demonstrate that the proposed network can achieve promising performance.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fast Local Representation learning with Adaptive Anchor Graph

Fast Local Representation Learning with Adaptive Anchor Grap...

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Canyu Zhang Feiping Nie Zheng Wang Rong Wang Xuelong Li School of Computer Science and Center for OPTical IMagery Analysis and Learning (OPTIMAL) Northwestern Polytechnical University Xi’an P. R. China School of Cybersecurity Northwestern Polytechnical University Xi’an Shaanxi P. R. China

ISBN: (纸本)9781728176055;9781728176062

Dimension reduction is an effective technology to embed data with high dimension to lower dimension space, where Linear Discriminant analysis (LDA), one of representative methods, only works with Gaussian distribution data. However, in order to solve non-Gaussian issue that only one cluster cannot well fit the distribution of same class, many graph-based discriminant analysis methods are proposed which capture local structure through measuring each pairwise distance. This is expense of time complexity because of the full-connections. In order to solve this issue, we propose a fast local representation learning with adaptive anchor graph to learn local structure information through similarity matrix in anchor-based graph. Notably, anchor points and similarity matrix are updated in subspace which is more precisely to capture local discriminant information. Experimental results on several synthetic and well-known datasets demonstrate the advantages of our method over the state-of-the-art methods.

关键词： Dimensionality reduction Conferences Signal processing algorithms Signal processing Gaussian distribution Iterative algorithms Linear discriminant analysis

来源：评论

学校读者我要写书评

暂无评论

Capped $\ell _p$-Norm LDA for Outliers Robust Dimension Reduction

引用

IEEE Signal Processing Letters 2020年 27卷 1315-1319页

作者： Zheng Wang Feiping Nie Canyu Zhang Rong Wang Xuelong Li School of Computer Science and Center for Optical Imagery Analysis and Learning Northwestern Polytechnical University Xi’an China School of Cybersecurity Northwestern Polytechnical University Xi’an China

Linear discriminant analysis technique is an effective strategy to solve the long-standing issue, i.e., the “curse of dimensionality” that brings many obstacles on high-dimensional data storage and analysis. However, the projections are prone to be affected, especially when the training set contains outlier samples whose distribution deviates from the globality. In many real-world applications, the outlier samples contaminated by noisy signal or spottiness have negative effects on the classification and clustering performance. To address this issue, we propose to develop a novel capped ℓ p -norm LDA model for robust dimension reduction against to outliers specifically. Proposed method integrates the capped ℓ p -norm based loss into the objective, which not only suppresses the light outliers but also works well even though the training set is contaminated seriously. Furthermore, we derive an alternative iterative re-weighted optimization algorithm to minimize the proposed objective based on capped ℓ p -norm with rigorous convergence proofs. Extensive experiments conducted on synthetic and real-world datasets demonstrate the robustness against to outliers of proposed method.

关键词： Robustness Signal processing algorithms Optimization Dimensionality reduction Training Iterative algorithms Convergence

来源：评论

学校读者我要写书评

暂无评论

Binary Classification with Supervised-like Biclustering and Adaboost

Binary Classification with Supervised-like Biclustering and ...

引用

International Conference on Information science and Control Engineering (ICISCE)

作者： Jianjun Sun Qinghua Huang School of Computer Science and Center for OPTical IMagery Analysis and Learning Northwestern Polytechnical University Xi’an Shaanxi People’s Republic of China School of Mechanical Engineering and Center for OPTical IMagery Analysis and Learning Northwestern Polytechnical University Xi’an Shaanxi People’s Republic of China

Biclustering based classification methods have achieved great success. In existing biclustering based classification methods the residue score threshold used for searching biclusters is fixed and the number of mined useful biclusters is small. To solve the two problems, in this study we proposed a novel binary classification method based on supervised-like biclustering and adaboost. Supervised-like means the biclusters is searched twice with two different bicluster quality indicators MSD and weights instead of unique indicator MSR in other biclustering based classification methods. In initial search, the residue score threshold is identical for all biclusters. In the second search, some biclusters of low quality are searched again with smaller residue threshold. Besides, due to some limitations, many biclusters cannot be found. To obtain more biclusters, two additional operations are adopted. In the proposed method, we initially mine column constant biclusters from datasets, then transform the biclusters to weak classifiers. Through adaboost, the initial strong binary classifier can be constructed. Finally, with supervised-like strategy, the better final strong binary classifier can be obtained. To verify the performance of the proposed method, it is compared with seven binary classification methods on two datasets. Experimental results demonstrated that the proposed method outperformed other binary classification methods.

关键词： Information science Control engineering Transforms Turning

来源：评论

学校读者我要写书评

暂无评论

Unsupervised semantic aggregation and deformable template matching for semi-supervised learning 20

Unsupervised semantic aggregation and deformable template ma...

引用

Proceedings of the 34th International Conference on Neural Information Processing Systems

作者： Tao Han Junyu Gao Yuan Yuan Qi Wang School of Computer Science and Center for OPTical IMagery Analysis and Learning Northwestern Polytechnical University Xi'an Shaanxi P.R. China

ISBN: (纸本)9781713829546

Unlabeled data learning has attracted considerable attention recently. However, it is still elusive to extract the expected high-level semantic feature with mere unsu-pervised learning. In the meantime, semi-supervised learning (SSL) demonstrates a promising future in leveraging few samples. In this paper, we combine both to propose an Unsupervised Semantic Aggregation and Deformable Template Matching (USADTM) framework for SSL, which strives to improve the classification performance with few labeled data and then reduce the cost in data annotating. Specifically, unsupervised semantic aggregation based on Triplet Mutual Information (T-MI) loss is explored to generate semantic labels for unlabeled data. Then the semantic labels are aligned to the actual class by the supervision of labeled data. Furthermore, a feature pool that stores the labeled samples is dynamically updated to assign proxy labels for unlabeled data, which are used as targets for cross-entropy minimization. Extensive experiments and analysis across four standard semi-supervised learning benchmarks validate that USADTM achieves top performance (e.g., 90.46% accuracy on CIFAR-10 with 40 labels and 95.20% accuracy with 250 labels).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Finite-time Consensus Tracking for Multi-motor System Based on Sliding Mode Control 6

Finite-time Consensus Tracking for Multi-motor System Based ...

引用

6th International Conference on Robotics and Automation Engineering, ICRAE 2021

作者： Zhang, Tao Yu, Dengxiu Wang, Zhen Jin, Xiaoyue Li, Shengjin Du, Feiping Northwestern Polytechnical University AVIC Qingan Group Co. Ltd School of Mechanical Engineering Xi'an China Northwestern Polytechnical University Unmanned System Research Institution Xi'an China Northwestern Polytechnical University Center for Optical Imagery Analysis and Learning Xi'an China Northwestern Polytechnical University School of Mechanical Engineering Xi'an China Xi'an Aerospace Propulsion Institute Science and Technology on Liquid Rocket Engine Laboratory Xi'an China

ISBN: (纸本)9781665406970

This study investigates a finite-time consensus tracking algorithm for synchronous coordination control of a multimotor system (MMS), which is extremely meaningful to some practical applications with the demand of high-precision performance and stringent convergence time. The graph theory is applied to realize the system modeling of the relationship between the internal parameters of MMS, which can reduce the complexity of the internal parameters of the system and make it more suitable for practical engineering. Then a distributed finite-time consensus controller based on sliding mode control for MMS is proposed, making the time costs in the reaching phase and sliding phase both finite. Simulations are given to illustrate the effectiveness of the proposed approach. © 2021 IEEE.

关键词： Sliding mode control

来源：评论

学校读者我要写书评

暂无评论

Video Frame Interpolation Via Residue Refinement

Video Frame Interpolation Via Residue Refinement

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Haopeng Li Yuan Yuan Qi Wang School of Computer Science and Center for OPTical IMagery Analysis and Learning (OPTIMAL) Northwestern Polytechnical University Xi’an P.R. China

ISBN: (数字)9781509066315

ISBN: (纸本)9781509066322

Video frame interpolation achieves temporal super-resolution by generating smooth transitions between frames. Although great success has been achieved by deep neural networks, the synthesized images stills suffer from poor visual appearance and unsatisfactory artifacts. In this paper, we propose a novel network structure that leverages residue refinement and adaptive weight to synthesize in-between frames. The residue refinement technique is used for optical flow and image generation for higher accuracy and better visual appearance, while the adaptive weight map combines the forward and backward warped frames to reduce the artifacts. Moreover, all submodules in our method are implemented by U-Net with less depths, so the efficiency is guaranteed. Experiments on public datasets demonstrate the effectiveness and superiority of our method over the state-of-the-art approaches.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Revisiting Fast Spectral Clustering with Anchor Graph

Revisiting Fast Spectral Clustering with Anchor Graph

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Cheng-Long Wang Feiping Nie Rong Wang Xuelong Li School of Computer Science and Center for OPTical IMagery Analysis and Learning (OPTIMAL) Northwestern Polytechnical University Xi’an P. R. China School of Cybersecurity and Center for OPTical IMagery Analysis and Learning (OPTIMAL) Northwestern Polytechnical University Xi’an P. R. China

ISBN: (数字)9781509066315

ISBN: (纸本)9781509066322

Many anchor-graph-based spectral clustering methods have been proposed to accelerate spectral clustering for large scale problems. In this paper, we revisit the popular large-scale spectral clustering method based on the anchor graph which is equivalent to the spectral decomposition on a similar matrix obtained using a second-order transition probability. However, due to the special structure of the bipartite graph, there is no stable distribution of the random walk process. The even-order transition probabilities may only a side view of the bipartite structure, resulting in breaking the independence of data points and leading to undesired artifacts for boundary samples. Therefore, we propose a Fast Spectral Clustering based on the Random Walk Laplacian (FRWL) method. The random walk Laplacian balances explicitly the popularity of anchors and the independence of data points, which keeps the structure of boundary samples. The experimental results demonstrate the efficiency and effectiveness of our method.

关键词：

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：