检索结果-南通市图书馆

Boosting Adversarial Training with Learnable Distribution

Computers, Materials & Continua 2024年第3期78卷 3247-3265页

作者： Kai Chen Jinwei Wang James Msughter Adeke Guangjie Liu yuewei dai School of Electronics and Information Engineering Nanjing University of Information Science and TechnologyNanjing210044China Key Laboratory of Intelligent Support Technology for Complex Environments Ministry of EducationNanjing210044China School of Computer and Software Nanjing University of Information Science and TechnologyNanjing210044China Nanjing Center for Applied Mathematics Nanjing211135China

In recent years,various adversarial defense methods have been proposed to improve the robustness of deep neural *** training is one of the most potent methods to defend against adversarial ***,the difference in the feature space between natural and adversarial examples hinders the accuracy and robustness of the model in adversarial *** paper proposes a learnable distribution adversarial training method,aiming to construct the same distribution for training data utilizing the Gaussian mixture *** distribution centroid is built to classify samples and constrain the distribution of the sample *** natural and adversarial examples are pushed to the same distribution centroid to improve the accuracy and robustness of the *** proposed method generates adversarial examples to close the distribution gap between the natural and adversarial examples through an attack algorithm explicitly designed for adversarial *** algorithm gradually increases the accuracy and robustness of the model by scaling ***,the proposed method outputs the predicted labels and the distance between the sample and the distribution *** distribution characteristics of the samples can be utilized to detect adversarial cases that can potentially evade the model *** effectiveness of the proposed method is demonstrated through comprehensive experiments.

关键词： Adversarial training feature space learnable distribution distribution centroid

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

基于UI-LSTM模型的短时降水预测研究

引用

无线电工程 2024年第1期54卷 47-54页

作者：包顺秦华旺戴跃伟陈浩然尹传豪南京信息工程大学电子与信息工程学院江苏南京210044 南京信息工程大学自动化学院江苏南京210044

降水临近预报是为了预测未来短时间的降雨量。现有大多数基于循环神经网络(Recurrent Neural Network,RNN)的降水预报模型,采用单一的卷积核对输入和隐藏状态的特征进行提取存在局部性,不能捕获雷达回波图中复杂的物理变化,且未有效提... 详细信息

降水临近预报是为了预测未来短时间的降雨量。现有大多数基于循环神经网络(Recurrent Neural Network,RNN)的降水预报模型,采用单一的卷积核对输入和隐藏状态的特征进行提取存在局部性,不能捕获雷达回波图中复杂的物理变化,且未有效提取时空相关性和对强降雨区域的精准预测。针对现有模型存在的问题,提出了UI-LSTM模型用于降水临近预报,能够有效地提取雷达回波序列的时空相关性,采用了U形结构,同时使用跳过连接进行特征拼接,学习到整个雷达回波图的上下文语义信息,且将浅层和深层信息进行特征融合。加入了Inception结构来代替ConvLSTM细胞结构中输入到输入和状态到状态的卷积,通过不同大小的卷积核,有效提取输入,隐藏状态的特征。在公开数据集(CIKM AnalytiCup 2017)进行实验并与其他模型进行对比实验。实验结果表明,UI-LSTM模型在HSS、CSI、MAE和SSIM指标整体上要远高于其他对比模型,且提高强降水天气预测的准确率。

关键词：降水临近预报循环神经网络特征融合 UI-LSTM Inception

来源：

维普期刊数据库博看期刊

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

基于D2D通信系统的无人机辅助物联网保密通信频谱共享方法

引用

无线电工程 2024年第3期54卷 695-703页

作者：丁徐飞田文刘光杰戴跃伟南京信息工程大学电子与信息工程学院江苏南京210044

保密通信作为物联网安全的一项核心技术近些年受到了广泛关注。无人机(UAV)由于其高机动和灵活部署等特性,被认为是提高物联网通信安全的有力手段。针对当前物联网中UAV辅助保密通信频谱利用率问题,研究了UAV和地面设备到设备(Device-to... 详细信息

保密通信作为物联网安全的一项核心技术近些年受到了广泛关注。无人机(UAV)由于其高机动和灵活部署等特性,被认为是提高物联网通信安全的有力手段。针对当前物联网中UAV辅助保密通信频谱利用率问题,研究了UAV和地面设备到设备(Device-to-Device, D2D)通信系统频谱共享的系统保密容量最大化,提出了一种基于块坐标下降(Block Coordinate Descent, BCD)法的频谱共享算法,该算法通过联合优化系统中发送方(UAV和地面设备对)的传输功率和无人机轨迹,在确保地面设备对和UAV能够共存的情况下最大化接收方的保密容量。具体来说,该算法通过差分凸规划(Difference of Two Convex Functions, DC)方法优化固定轨迹下的系统功率,利用连续凸优化方法解决给定发射功率的UAV轨迹优化问题。仿真结果表明,所提方法在多种参数配置下都能有效提高系统的保密容量,并且和其他算法相比有更低的功率消耗。

关键词：无人机保密容量终端设备直连通信系统凸优化

来源：

维普期刊数据库

同方期刊数据库博看期刊

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Detecting JPEG image forgery based on double compression

引用

Journal of Systems Engineering and Electronics 2009年第5期20卷 1096-1103页

作者： Wang Junwen Liu Guangjie dai yuewei Wang Zhiquan School of Automation Nanjing Univ. of Science &: Technology Nanjing 210094 P. R. China

Detecting the forgery parts from a double compressed image is very important and urgent work for blind authentication. A very simple and efficient method for accomplishing the task is proposed. Firstly, the probabilistic model with periodic effects in double quantization is analyzed, and the probability of quantized DCT coefficients in each block is calculated over the entire iraage. Secondly, the posteriori probability of each block is computed according to Bayesian theory and the results mentioned in first part. Then the mean and variance of the posteriori probability are to be used for judging whether the target block is tampered. Finally, the mathematical morphology operations are performed to reduce the false alarm probability. Experimental results show that the method can exactly locate the doctored part, and through the experiment it is also found that for detecting the tampered regions, the higher the second compression quality is, the more exact the detection efficiency is.

关键词： image forgery JPEG double compression quantization posteriori probability.

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Improved quantum bacterial foraging algorithm for tuning parameters of fractional-order PID controller

引用

Journal of Systems Engineering and Electronics 2018年第1期29卷 166-175页

作者： LIU Lu SHAN Liang dai yuewei LIU Chenglin QI Zhidong School of Automation Nanjing University of Science and Technology Key Laboratory of Advanced Process Control for Light Industry (Ministry of Education) Institute of AutomationJiangnan University

The quantum bacterial foraging optimization(QBFO)algorithm has the characteristics of strong robustness and global searching ability. In the classical QBFO algorithm, the rotation angle updated by the rotation gate is discrete and constant,which cannot affect the situation of the solution space and limit the diversity of bacterial population. In this paper, an improved QBFO(IQBFO) algorithm is proposed, which can adaptively make the quantum rotation angle continuously updated and enhance the global search ability. In the initialization process, the modified probability of the optimal rotation angle is introduced to avoid the existence of invariant solutions. The modified operator of probability amplitude is adopted to further increase the population *** tests based on benchmark functions verify the effectiveness of the proposed algorithm. Moreover, compared with the integerorder PID controller, the fractional-order proportion integration differentiation(PID) controller increases the complexity of the system with better flexibility and robustness. Thus the fractional-order PID controller is applied to the servo system. The tuning results of PID parameters of the fractional-order servo system show that the proposed algorithm has a good performance in tuning the PID parameters of the fractional-order servo system.

关键词： bacterial foraging algorithm fractional-order quantum rotation gate proportion integration differentiation(PID) servo system

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

PHOTOREALISTIC COMPUTER GRAPHICS FORENSICS BASED ON LEADING DIGIT LAW

引用

Journal of Electronics(China) 2011年第1期28卷 95-100页

作者： Xu Bo Wang Junwen Liu Guangjie dai yuewei School of Automation Nanjing University of Science ＆ Technology Nanjing 210004 China

As the advent and growing popularity of image rendering software,photorealistic computer graphics are becoming more and more perceptually indistinguishable from photographic *** the faked images are abused,it may lead to potential social,legal or private *** this end,it is very necessary and also challenging to find effective methods to differentiate between *** this paper,a novel leading digit law,also called Benford's law,based method to identify computer graphics is *** specifically,statistics of the most significant digits are extracted from image's Discrete Cosine Transform(DCT) coefficients and magnitudes of image's gradient,and then the Support Vector Machine(SVM) based classifiers are *** of experiments on the image datasets indicate that the proposed method is comparable to prior ***,it possesses low dimensional features and low computational complexity.

关键词： Leading digit law Benford’s law Digital image forensic Computer graphics

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

On fast estimation of direction of arrival for underwater acoustic target based on sparse Bayesian learning

引用

Chinese Journal of Acoustics 2017年第1期36卷 102-112页

作者： WANG Biao ZHU Zhihui dai yuewei School of Electronic and Information Jiangsu University of Science and Technology Zhenjiang 212003 Key Laboratory of Underwater Acoustic Warfare Technology Beijing 100094

The Direction of Arrival （DOA） estimation methods for underwater acoustic target using Temporally Multiple Sparse Bayesian Learning （TMSBL） as the reconstructing algorithm have the disadvantage of slow computing speed. To solve this problem, a fast underwater acoustic target direction of arrival estimation was proposed. Analyzing the model characteristics of block-sparse Bayesian learning framework for DOA estimation, an algorithm was proposed to obtain the value of core hyper-parameter through MacKay＇s fixed-point method to estimate the DOA. By this process, it will spend less time for computation and provide more superior recovery performance than TMSBL algorithm. Simulation results verified the feasibility and effectiveness of the proposed algorithm.

关键词： On fast estimation of direction of arrival for underwater acoustic target based on sparse Bayesian learning DOA

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Sliding Mode Control of T-Shaped Pedestrian Channel

引用

Journal of Shanghai Jiaotong university(Science) 2020年第4期25卷 478-485页

作者：常路单梁李军戴跃伟 School of Automation Nanjing University of Science and TechnologyNanjing 210094China

The study of pedestrian evacuation in channels with different structures is important among optimizing through efficiency,avoiding accidents and designing *** is of significant reference to set up a specific model on account of the real traffic system and design an appropriate controller to apply on *** paper establishes a macroscopic model for T-shaped channel with a number of controlled entrances basing on the law of mass ***,with the method of cascade,a kind of sliding mode controller is designed to achieve the control target of avoiding the blocking and maximizing the pedestrian flow of the whole model,the stability of the system is proved by the Lyapunov stability theory,and the boundary layer method is applied to restrain the chattering of the ***,simulation results show the efficiency of the sliding mode controller and the improvement brought by the boundary layer method.

关键词： nonlinear stability analysis sliding mode control macroscopic model boundary layer method

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

面向智能运维的KPI异常检测模型研究

引用

重庆理工大学学报（自然科学） 2022年第6期36卷 181-188页

作者：陈倩戴跃伟刘光杰南京信息工程大学电子与信息工程学院南京210044

目前大多数KPI异常检测方法无法提取数据的有效特征,且未充分考虑KPI的时间相关性,检测不够准确。因此,针对KPI序列的异常检测问题,提出了一种基于变分自编码器(VAE)结合门控循环单元网络(GRU)的异常检测方法。对KPI序列进行预处理,使... 详细信息

目前大多数KPI异常检测方法无法提取数据的有效特征,且未充分考虑KPI的时间相关性,检测不够准确。因此,针对KPI序列的异常检测问题,提出了一种基于变分自编码器(VAE)结合门控循环单元网络(GRU)的异常检测方法。对KPI序列进行预处理,使用滑动窗口将序列输入模型,通过变分自编码器捕获KPI序列的有效特征,并基于这些特征建立门控循环单元网络进行预测,再输入解码器重构序列,在此基础上利用计算得到的重构误差与给定阈值对比来判断异常。实验结果表明,与单独使用变分自编码器和门控循环单元网络进行异常检测相比,本模型在准确检测KPI异常点方面表现出较好的优势。

关键词： KPI异常检测变分自编码器门控循环单元

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

时间反转滤波器组多载波水声通信方法

引用

声学学报 2020年第1期45卷 38-44页

作者：王彪方涛戴跃伟江苏科技大学电子信息学院镇江212003 水声对抗技术重点实验室北京100094

针对水声滤波器组多载波(Filter Bank MultiCarrier,FBMC)传统均衡方法效率低、复杂度高等问题,提出了单阵元无源时间反转镜(Passive Time Reversal Mirror,PTRM)和虚拟时间反转镜(Virtual Time Reversal Mirror,VTRM)的均衡算法,该算法利用时间聚焦原理可以降低水声信道带来的符号间干扰(Inter Symbol Interference,ISI),并且不需要在FBMC符号中插入导频,进一步提高了通信速率.通过仿真和外场实验验证了水声FBMC系统采用单阵元PTRM和VTRM的有效性,结果表明,这两种均衡方法能够有效减少ISI,提高系统可靠性能。

关键词：水声通信时间反转镜滤波器组多载波水声信道

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

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：