检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Chen, Jian Zhang, Xuxin Zhang, Rui Wang, Chen Liu, Ling School of Electronic Information and Communications Huazhong University of Science and Technology Wuhan430074 China Hubei Key Laboratory of Transportation Internet of Things School of Computer Science and Technology Wuhan University of Technology Wuhan430070 China College of Computing Georgia Institute of Technology AtlantaGA30332-0765 United States

Machine learning techniques have been widely applied to various applications. However, they are potentially vulnerable to data poisoning attacks, where sophisticated attackers can disrupt the learning procedure by injecting a fraction of malicious samples into the training dataset. Existing defense techniques against poisoning attacks are largely attack-specific: they are designed for one specific type of attacks but do not work for other types, mainly due to the distinct principles they follow. Yet few general defense strategies have been developed. In this paper, we propose De-Pois, an attack-agnostic defense against poisoning attacks. The key idea of De-Pois is to train a mimic model the purpose of which is to imitate the behavior of the target model trained by clean samples. We take advantage of Generative Adversarial Networks (GANs) to facilitate informative training data augmentation as well as the mimic model construction. By comparing the prediction differences between the mimic model and the target model, De-Pois is thus able to distinguish the poisoned samples from clean ones, without explicit knowledge of any ML algorithms or types of poisoning attacks. We implement four types of poisoning attacks and evaluate De-Pois with five typical defense methods on different realistic datasets. The results demonstrate that De-Pois is effective and efficient for detecting poisoned data against all the four types of poisoning attacks, with both the accuracy and F1-score over 0.9 on average. Copyright © 2021, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Research on Infrared Image Processing Algorithm of Generator Carbon Brush Based on Adaptive Stochastic Resonance Array

Research on Infrared Image Processing Algorithm of Generator...

引用

2021 China Automation Congress, CAC 2021

作者： Jiao, Shangbin Huang, Weichao Geng, Bo Zhang, Youmin Wu, Xiaohui Wu, Pengyue Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xi 'an University of Technology Xi'an710048 China Department of Mechanical Industrial and Aerospace Engineering Concordia University MontrealQCH3G 1M8 Canada Huaneng Weihai Power Generation Co. Ltd Weihai264200 China Xi'an Thermal Power Research Institute Co. Ltd. Xi'an710048 China

ISBN: (纸本)9781665426473

The safe and stable operation of the generator set is related to the national economy and people's livelihood. As an important part of generator excitation system, carbon brush and slip ring temperature monitoring can effectively evaluate the state of the generator, which plays a vital role. Most of the current researches use infrared images to monitor carbon brush temperature, but there are often problems such as high noise and unclear focus. According to the requirements of infrared image display, this paper combines the method of bistable stochastic resonance to denoise the image, and on this basis, introduces the adaptive stochastic resonance array method to denoise the gray image. Experimental results show that compared with common image denoising methods and classic bistable stochastic resonance, the adaptive stochastic resonance array method has improved both visual effects and peak signal-to-noise ratio (PSNR), which further proves the good application of stochastic resonance in weak signal detection and extraction. © 2021 IEEE

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Three-Dimensional Charge Density Wave and Surface-Dependent Vortex-Core States in a Kagome Superconductor CsV3Sb5

引用

Physical Review X 2021年第3期11卷 031026-031026页

作者： Zuowei Liang Xingyuan Hou Fan Zhang Wanru Ma Ping Wu Zongyuan Zhang Fanghang Yu J.-J. Ying Kun Jiang Lei Shan Zhenyu Wang X.-H. Chen Department of Physics and Chinese Academy of Sciences Key laboratory of Strongly-coupled Quantum Matter Physics University of Science and Technology of China Hefei Anhui 230026 China Information Materials and Intelligent Sensing Laboratory of Anhui Province Institutes of Physical Science and Information Technology Anhui University Hefei 230601 China Beijing National Laboratory for Condensed Matter Physics and Institute of Physics Chinese Academy of Sciences Beijing 100190 China Key Laboratory of Structure and Functional Regulation of Hybrid Materials of Ministry of Education Anhui University Hefei 230601 China CAS Center for Excellence in Quantum Information and Quantum Physics Hefei Anhui 230026 China Collaborative Innovation Center of Advanced Microstructures Nanjing 210093 China

The transition-metal-based kagome metals provide a versatile platform for correlated topological phases hosting various electronic instabilities. While superconductivity is rare in layered kagome compounds, its interplay with nontrivial topology could offer an engaging space to realize exotic excitations of quasiparticles. Here, we use scanning tunneling microscopy to study a newly discovered Z2 topological kagome metal CsV3Sb5 with a superconducting ground state. We observe charge modulation associated with the opening of an energy gap near the Fermi level. When across single-unit-cell surface step edges, the intensity of this charge modulation exhibits a π-phase shift, suggesting a three-dimensional 2×2×2 charge density wave ordering. Interestingly, while conventional Caroli–de Gennes–Matricon bound states are observed inside the superconducting vortex on the Sb surfaces, a robust zero-bias conductance peak emerges that does not split in a large distance when moving away from the vortex center on the Cs 2×2 surfaces, resembling the Majorana bound states arising from the superconducting Dirac surface states in Bi2Te3/NbSe2 heterostructures. Our findings establish CsV3Sb5 as a promising candidate for realizing exotic excitations at the confluence of nontrivial lattice geometry, topology and multiple electronic orders.

关键词： Charge density waves Local density of states Majorana bound states Topological materials Topological superconductors

来源：评论

学校读者我要写书评

暂无评论

Two-Way Feature-Aligned And Attention-Rectified Adversarial Training

Two-Way Feature-Aligned And Attention-Rectified Adversarial ...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Haitao Zhang Fan Jia Quanxin Zhang Yahong Han Xiaohui Kuang Yu-an Tan College of Intelligence and Computing Tianjin University Tianjin China Beijing Institute of Technology National Key Laboratory of Science and Technology on Information System Security

ISBN: (数字)9781728113319

ISBN: (纸本)9781728113326

Adversarial training increases robustness by augmenting training data with adversarial examples. However, vanilla adversarial training may be overfitting to certain adversarial attacks. Small perturbations in images bring in error which is gradually amplified when forwarded through the model so that the error leads to wrong classification. Besides, small perturbations will also distract classifier's attention to significant features that are relevant to the true label. In this paper, we propose a novel two-way feature-aligned and attention-rectified adversarial training (FAAR) to improve adversarial training (AT). FAAR utilizes two-way feature alignment and attention rectification to mitigate the problems mentioned above. FAAR effectively suppresses perturbations in lowlevel, high-level and global features by moving features of perturbed images towards those of clean images with twoway feature alignment. It also leads the model into focusing more on useful features which are correlated with true label through rectifying gradient-weighted attention. Besides, feature alignment activates attention rectification by reducing perturbations in high-level feature. Our proposed method FAAR surpasses other existing AT methods in three aspects. First, it pushes the model to keep invariant when dealing with different adversarial attacks and different magnitude of perturbations. Second, it can be applied to any convolution neural networks. Third, the training process is end-to-end. For experiments, FAAR shows promising defense performance on CIFAR-10 and ImageNet.

关键词： Training Perturbation methods Robustness Mathematical model Convolution Image color analysis Computational modeling

来源：评论

学校读者我要写书评

暂无评论

$run$runData: Re-Distributing Data via Piggybacking for Geo-Distributed Data Analytics Over Edges

引用

IEEE Transactions on Parallel and Distributed Systems 2021年第1期33卷 40-55页

作者： Yibo Jin Zhuzhong Qian Song Guo Sheng Zhang Lei Jiao Sanglu Lu Department of Computer Science and Technology State Key Laboratory for Novel Software Technology Nanjing University Nanjing Jiangsu China Department of Computing The Hong Kong Polytechnic University Hung Hom Hong Kong Hong Kong Polytechnic University Shenzhen Research Institute Shenzhen Guangdong China Department of Computer and Information Science University of Oregon Eugene OR USA

Efficiently analyzing geo-distributed datasets is emerging as a major demand in a cloud-edge system. Since the datasets are often generated in closer proximity to end users, traditional works mainly focus on offloading proper tasks from those hotspot edges to the datacenter to decrease the overall completion time of submitted jobs in a one-shot manner. However, optimizing the completion time of current job alone is insufficient in a long-term scope since some datasets would be used multiple times. Instead, optimizing the data distribution is much more efficient and could directly benefit forthcoming jobs, although it may postpone the execution of current one. Unfortunately, due to the throwaway feature of data fetcher, existing data analytics systems fail to re-distribute corresponding data out of hotspot edges after the execution of data analytics. In order to minimize the overall completion time for a sequence of jobs as well as to guarantee the performance of current one, we propose to re-distribute the data along with task offloading, and formulate corresponding ε-bounded data-driven task scheduling problem over wide area network under the consideration of edge heterogeneity. We design an online schema run Data, which offloads proper tasks and related data via piggybacking to the datacenter based on delicately calculated probabilities. Through rigorous theoretical analysis, run Data is proved concentrated on its optimum with high probability. We implement run Data based on Spark and HDFS. Both testbed results and trace-driven simulations show that run Data re-distributes proper data via piggybacking and achieves up to 37 percent reduction on average response time compared with state-of-the-art schemas.

关键词： Task analysis Data analysis Wide area networks Data models Servers Optimization Videos

来源：评论

学校读者我要写书评

暂无评论

Supervised learning for robust quantum control in composite-pulse systems

arXiv

引用

arXiv 2023年

作者： Shi, Zhi-Cheng Ding, Jun-Tong Chen, Ye-Hong Song, Jie Xia, Yan Yi, X.X. Nori, Franco Fujian Key Laboratory of Quantum Information and Quantum Optics Fuzhou University Fuzhou350108 China Department of Physics Fuzhou University Fuzhou350108 China Theoretical Quantum Physics Laboratory Cluster for Pioneering Research RIKEN Saitama Wakoshi351-0198 Japan Quantum Information Physics Theory Research Team Center for Quantum Computing RIKEN Saitama Wakoshi351-0198 Japan Department of Physics Harbin Institute of Technology Harbin150001 China Center for Quantum Sciences School of Physics Northeast Normal University Changchun130024 China Department of Physics University of Michigan Ann ArborMI48109-1040 United States

In this work, we develop a supervised learning model for implementing robust quantum control in composite-pulse systems, where the training parameters can be either phases, detunings, or Rabi frequencies. This model exhibits great resistance to all kinds of systematic errors, including single, multiple, and time-varying errors. We propose a modified gradient descent algorithm for adapting the training of phase parameters, and show that different sampling methods result in different robust performances. In particular, there is a trade-off between high fidelity and robustness for a given number of training parameters, and both can be simultaneously enhanced by increasing the number of training parameters (pulses). For its applications, we demonstrate that the current model can be used for achieving high-fidelity arbitrary superposition states and universal quantum gates in a robust manner. This work provides a highly efficient learning model for fault-tolerant quantum computation by training various physical parameters. © 2023, CC0.

关键词： Systematic errors

来源：评论

学校读者我要写书评

暂无评论

Associating multi-scale receptive fields for fine-grained recognition

arXiv

引用

arXiv 2020年

作者： Ye, Zihan Hu, Fuyuan Liu, Yin Xia, Zhenping Lyu, Fan Liu, Pengqing Suzhou University of Science and Technology Virtual Reality Key Laboratory of Intelligent Interaction and Application Technology of Suzhou Suzhou Key Laboratory for Big Data and Information Service Shanghai Institute of Technology Tianjin University

Extracting and fusing part features have become the key of fined-grained image recognition. Recently, Non-local (NL) module has shown excellent improvement in image recognition. However, it lacks the mechanism to model the interactions between multi-scale part features, which is vital for fine-grained recognition. In this paper, we propose a novel cross-layer non-local (CNL) module to associate multi-scale receptive fields by two operations. First, CNL computes correlations between features of a query layer and all response layers. Second, all response features are weighted according to the correlations and are added to the query features. Due to the interactions of cross-layer features, our model builds spatial dependencies among multi-level layers and learns more discriminative features. In addition, we can reduce the aggregation cost if we set low-dimensional deep layer as query layer. Experiments are conducted to show our model achieves or surpasses state-of-the-art results on three benchmark datasets of fine-grained classification. Our codes can be found at ***/FouriYe/CNL-ICIP2020. Copyright © 2020, The Authors. All rights reserved.

关键词： Image recognition

来源：评论

学校读者我要写书评

暂无评论

A dual-band rectifier with improved T-section matching network for RF energy harvester

A dual-band rectifier with improved T-section matching netwo...

引用

International Conference on Microwave and Millimeter Wave technology Proceedings

作者： Baohong Liu Yong Qin Ying Chen Runtian Wei Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing Nanchang Institute of Technology Nanchang China Nanchang Institute of Technology Nanchang China

ISBN: (数字)9781728157337

ISBN: (纸本)9781728157344

In this paper, a dual-band rectifier with simple matching network composed of improved T-section. Diode's impedances at two frequencies are turned into appropriate values for physical realization by the additional stub and the rest parts make them be turned into the conjugated pairs. A dual-band rectifier operating at 0.915-GHz and 2.45-GHz is implemented and simulated for verification. Simulated results show the proposed rectifier can operate dual bands with power conversion efficiency of 75% and 65% at each frequency. The proposed matching network rectifier is very suitable for RF energy harvester due to convenience and simplicity.

关键词： Radio frequency Microwave technology Millimeter wave technology Rectifiers Dual band Proposals Power conversion

来源：评论

学校读者我要写书评

暂无评论

Pan-sharpening via Dual Latent Variable Constraint Network

引用

information Fusion 2025年 124卷

作者： Jiaming Wang Dongyue Luo Xiao Huang Ruiqian Zhang Yu Wang Tao Lu Hubei Key Laboratory of Intelligent Robot Wuhan Institute of Technology Wuhan 430205 China The Department of Environmental Sciences Emory University Atlanta 30322 USA Institute of Photogrammetry and Remote Sensing Chinese Academy of Surveying and Mapping Beijing 100036 China State Key Laboratory for Information Engineering in Surveying Mapping and Remote Sensing Wuhan University Wuhan 430079 China

Most existing pan-sharpening methods usually achieve fusion by directly injecting the information of panchromatic (PAN) images into multi-spectral (MS) images in a strongly coupled manner. Although this approach can enhance spatial details, it ignores the differences and complementarities between the two at the feature level, which can easily lead to spectral distortion and loss of spatial texture information. To address these challenges, we present a latent variable constrained network composed of two main components: a quantized variational autoencoder (QVA) and a dual conditional fusion (DCF) network, which jointly optimize spectral fidelity and spatial detail reconstruction. The proposed framework is trained in two sequential stages. First, a PAN encoder–decoder is established via a variational autoencoder-based model, thereby acting as a multi-scale latent variable extractor for the PAN domain. Next, we introduce a double-constraint invertible block, which integrates the multi-scale PAN latent variables and MS features under a low-coupling fusion scheme. This design significantly enhances both spectral accuracy and spatial detail recovery. Extensive experiments on multiple satellite image datasets validate the superiority of the proposed method. The results demonstrate that our approach consistently surpasses state-of-the-art pan-sharpening techniques in terms of both subjective visual evaluations and objective performance metrics. The source code will be made available at: https://***/jiaming-wang/Project .

关键词： Conditional learning Multi-scale fusion Pan-sharpening Variational autoencoders

来源：评论

学校读者我要写书评

暂无评论

Instructing Text-to-Image Diffusion Models via Classifier-Guided Semantic Optimization

arXiv

引用

arXiv 2025年

作者： Chang, Yuanyuan Yao, Yinghua Qin, Tao Wang, Mengmeng Tsang, Ivor Dai, Guang MOE Key Laboratory for Intelligent Networks and Network Security Xi’an Jiaotong University China Center for Frontier AI Research Agency for Science Technology and Research Singapore Institute of High Performance Computing Agency for Science Technology and Research Singapore Zhejiang University of Technology China SGIT AI Lab State Grid Corporation of China China

Text-to-image diffusion models have emerged as powerful tools for high-quality image generation and editing. Many existing approaches rely on text prompts as editing guidance. However, these methods are constrained by the need for manual prompt crafting, which can be time-consuming, introduce irrelevant details, and significantly limit editing performance. In this work, we propose optimizing semantic embeddings guided by attribute classifiers to steer text-to-image models toward desired edits, without relying on text prompts or requiring any training or fine-tuning of the diffusion model. We utilize classifiers to learn precise semantic embeddings at the dataset level. The learned embeddings are theoretically justified as the optimal representation of attribute semantics, enabling disentangled and accurate edits. Experiments further demonstrate that our method achieves high levels of disentanglement and strong generalization across different domains of data. Code is available at https://***/Chang-yuanyuan/CASO. Copyright © 2025, The Authors. All rights reserved.

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：