Vehicle Abnormal Behavior Detection Based on Dense Block and Soft Thresholding

With the rapid advancement of social economies,intelligent transportation systems are gaining increasing atten-tion.Central to these systems is the detection of abnormal vehicle behavior,which remains a critical challenge due to the complexity of urban roadways and the variability of external conditions.Current research on detecting abnormal traffic behaviors is still nascent,with significant room for improvement in recognition accuracy.To address this,this research has developed a new model for recognizing abnormal traffic behaviors.This model employs the R3D network as its core architecture,incorporating a dense block to facilitate feature reuse.This approach not only enhances performance with fewer parameters and reduced computational demands but also allows for the acquisition of new features while simplifying the overall network structure.Additionally,this research integrates a self-attentive method that dynamically adjusts to the prevailing traffic conditions,optimizing the relevance of features for the task at hand.For temporal analysis,a Bi-LSTM layer is utilized to extract and learn from time-based data nuances.This research conducted a series of comparative experiments using the UCF-Crime dataset,achieving a notable accuracy of 89.30%on our test set.Our results demonstrate that our model not only operates with fewer parameters but also achieves superior recognition accuracy compared to previous models.

PROJECTED GRADIENT DESCENT BASED ON SOFT THRESHOLDING IN MATRIX COMPLETION

Matrix completion is the extension of compressed sensing.In compressed sensing,we solve the underdetermined equations using sparsity prior of the unknown signals.However,in matrix completion,we solve the underdetermined equations based on sparsity prior in singular values set of the unknown matrix,which also calls low-rank prior of the unknown matrix.This paper firstly introduces basic concept of matrix completion,analyses the matrix suitably used in matrix completion,and shows that such matrix should satisfy two conditions:low rank and incoherence property.Then the paper provides three reconstruction algorithms commonly used in matrix completion:singular value thresholding algorithm,singular value projection,and atomic decomposition for minimum rank approximation,puts forward their shortcoming to know the rank of original matrix.The Projected Gradient Descent based on Soft Thresholding(STPGD),proposed in this paper predicts the rank of unknown matrix using soft thresholding,and iteratives based on projected gradient descent,thus it could estimate the rank of unknown matrix exactly with low computational complexity,this is verified by numerical experiments.We also analyze the convergence and computational complexity of the STPGD algorithm,point out this algorithm is guaranteed to converge,and analyse the number of iterations needed to reach reconstruction error.Compared the computational complexity of the STPGD algorithm to other algorithms,we draw the conclusion that the STPGD algorithm not only reduces the computational complexity,but also improves the precision of the reconstruction solution.

Parameter Estimation of Multiple Frequency-Hopping Signals Based on Space-Time-Frequency Analysis by Atomic Norm Soft Thresholding with Missing Observations

In this paper,we address the problem of multiple frequency-hopping(FH)signal parameters estimation in the presence of random missing observations.A space-time matrix with random missing observations is acquired by a uniform linear array(ULA).We exploit the inherent incomplete data processing capability of atomic norm soft thresholding(AST)to analyze the space-time matrix and complete the accurate estimation of the hopping time and frequency of the received FH signals.The hopping time is obtained by the sudden changes of the spatial information,which is implemented as the boundary to divide the time domain signal so that each segment of the signal is a superposition of time-invariant multiple components.Then,the frequency of multiple signal components can be estimated precisely by AST within each segment.After obtaining the above two parameters of the hopping time and the frequency of signals,the direction of arrival(DOA)can be directly calculated by them,and the network sorting can be realized.Results of simulation show that the proposed method is superior to the existing technology.Even when a large portion of data observations is missing,as the number of array elements increases,the proposed method still achieves acceptable accuracy of multi-FH signal parameters estimation.

Research on fiber optic gyro signal de-noising based on wavelet packet soft-threshold

Gyro＇s drift is not only the main drift error which influences gyro＇s precision but also the primary factor that affects gyro＇s reliability. Reducing zero drift and random drift is a key problem to the output of a gyro signal. A three-layer de-nosing threshold algorithm is proposed based on the wavelet decomposition to dispose the signal which is collected from a running fiber optic gyro （FOG）. The coefficients are obtained from the three-layer wavelet packet decomposition. By setting the high frequency part which is greater than wavelet packet threshold as zero, then reconstructing the nodes which have been filtered out noise and interruption, the soft threshold function is constructed by the coefficients of the third nodes. Compared wavelet packet de-noise with forced de-noising method, the proposed method is more effective. Simulation results show that the random drift compensation is enhanced by 13.1%, and reduces zero drift by 0.052 6°/h.

A Model-Based Soft Decision Approach for Speech Enhancement

Many speech enhancement algorithms that deal with noise reduction are based on a binary masking decision(termed as the hard decision), which may cause some regions of the synthesized speech to be discarded. In view of the problem, a soft decision is often used as an optimal technique for speech restoration. In this paper, considering a new fashion of speech and noise models, we present two model-based soft decision techniques. One technique estimates a ratio mask generated by the exact Bayesian estimators of speech and noise. For the second technique, we consider one issue that an optimum local criterion(LC) for a certain SNR may not be appropriate for other SNRs. So we estimate a probabilistic mask with a variable LC. Experimental results show that the proposed method achieves a better performance than reference methods in speech quality.

Soft Handover Probability Determination Considering New Direction of Motion

This paper presents prior determination of soft handover probability considering new direction of motion of mobile station (MS) coinciding with gravitation point of cells. Our simulation results for 3-cell scenario and considered new direction of MS motion can be potentially used as advance input to soft handover algorithms to minimize number of handovers.

WiCare:一种非接触式的老人如厕跌倒监测模型

老人在卫生间内的跌倒行为存在因救助及时性差而导致严重危害的风险,因此高效快捷的如厕跌倒监测研究具有重要意义。针对当前基于Wi-Fi感知的跌倒监测方法中存在的受噪声影响大而特征提取不充分、监测精度有限的问题,提出了一种基于多级离散小波变换和软阈值处理的信号降噪算法,及一种融合卷积神经网络、双向长短期记忆网络及自注意力机制的非接触式如厕跌倒监测模型WiCare。首先,从原始CSI数据中提取振幅作为基础数据;其次,使用多级离散小波变换和软阈值处理进行感知数据降噪;然后,将感知数据进行多维重构,以更准确地表征跌倒行为特征;最后,利用WiCare提取感知数据中的有效特征,进而实现卫生间如厕跌倒行为监测功能。实验结果表明,WiCare在居家卫生间环境下对跌倒行为监测的准确率为99.41%,与其他同类模型相比,WiCare的识别准确率高,模型复杂度低,且泛化能力更强。

基于多残差注意力深度收缩网络的超微光图像增强方法

超微光成像可在极度黑暗的环境中给观察者提供近乎白昼的视觉体验,在许多民用和军事应用中起着至关重要的作用。超微光环境下拍摄的图像和视频通常存在亮度与对比度极低、噪声水平高、场景细节和色彩严重缺失等固有缺陷,近年来,深度学习为超微光成像的研究带来了新的机遇。文中采集并提供了一组实用性更强的超微光训练数据集,提出了一种多残差注意力深度收缩网络(Multi Residual Attention Shrinkage Network),以此实现了一种新的超微光成像方法。通过成功研制的小型化样机证实了该方法的工业量产前景。实现了基于通道注意力和空间注意力的残差内注意力机制,以及基于深度软阈值收缩的外注意力机制,不仅可以有效提取并还原极低照度环境下的图像细节信息,恢复场景真实色彩,而且可以有效去除此类环境下由成像设备感光不足带来的巨量噪声。实测效果显示该方法可对极低照度环境进行有效的增强且实时性高。通过与多种业界最新方法比较,文中方法在主观视觉体验以及客观参数两方面均表现更好。

基于多尺度融合和时空特征的网络入侵检测模型

针对入侵检测模型提取特征能力不足,且流量数据中含冗余噪声的问题,提出一种基于多尺度融合和时空特征的ML-PFN入侵检测模型。采用多尺度特征融合技术分别提取数据中浅层特征信息和深层特征信息,使模型学习的特征更加丰富;采用软阈值函数和注意力机制自动选择合适的阈值,减少噪声及不相关信息对模型的干扰;融合时空特征构成多尺度空间特征提取长短时记忆-并行特征网络(MSFE LSTM-parallel feature network, ML-PFN)模型,并应用于网络入侵检测。通过3个公开数据集进行性能评估,实验结果表明,ML-PFN模型对比其它5种分类模型各项指标效果最好,在训练时长适中的同时准确率达到96.45%。

An alternative 2D ICEEMDAN-based denoising method and its application in processing magnetic anomaly data

Due to environmental noise and human factors,magnetic data collected in the field often contain various noises and interferences that significantly affect the subsequent data processing and interpretation.Empirical Mode Decomposition(EMD),an adaptive multiscale analysis method for nonlinear and non-stationary signals,is widely used in geophysical and geodetic data processing.Compared with traditional EMD,Improved Complete Ensemble EMD with Adaptive Noise(ICEEMDAN)is more effective in addressing the problem of mode mixing.Based on the principles of 1D ICEEMDAN,this paper presents an alternative algorithm for 2D ICEEMDAN,extending its application to two-dimensional scenarios.The effectiveness of the proposed approach is demonstrated through synthetic signal experiments,which show that the 2D ICEEMDAN exhibits a weaker mode mixing effect compared to the traditional bidimensional EMD(BEMD)method.Furthermore,to improve the performance of the denoising method based on 2D ICEEMDAN and preserve useful signals in high-frequency components,an improved soft thresholding technique is introduced.Synthetic magnetic anomaly data testing indicates that our denoising method effectively preserves signal continuity and outperforms traditional soft thresholding methods.To validate the practical application of this improved threshold denoising method based on 2D ICEEMDAN,it is applied to ground magnetic survey data in the Yandun area of Xinjiang.The results demonstrate the effectiveness of the method in removing noise while retaining essential information from practical magnetic anomaly data.In particular,practical applications suggest that 2D ICEEMDAN can extract trend signals more accurately than the BEMD.In conclusion,as a potential tool for multi-scale decomposition,the 2D ICEEMDAN is versatile in processing and analyzing 2D geophysical and geodetic data.

基于双阶段特征提取网络的ECG降噪分类算法

临床采集到的标准12导联心电图常含有噪声,影响了心电信号分类结果的准确度,为此提出了一种基于双阶段特征提取网络的心电图(ECG)降噪分类算法。首先,在空间特征提取阶段,由深度耦合软阈值化去噪方法的残差收缩网络从输入的12导联标准心电信号中提取空间特征;其次,在时间特征提取阶段,由长短期记忆网络与注意力机制结合继续从心电信号中提取时间特征;最后,通过全连接网络层融合提取到的空间特征与时间特征,输出9个类别的概率预测分布。在CPSC2018数据集上与其他同类型先进分类算法进行了对比实验,验证所提算法的效果,实验结果表明:提出的分类算法在对9类ECG信号进行分类时平均F1分数达到0.854,在各项指标上表现更优。此外,实验证明所提算法在含噪数据中的表现也优于其他主流网络,充分证明了所提算法对于含噪心电信号的降噪分类性能,该算法也可应用于其他类似含噪声生理信号的分析和处理。

DRSN与集成融合的OFDM辐射源个体识别方法

针对在低信噪下通信辐射源识别率低的问题,提出一种DRSN(Deep Residual Shrinkage Networks)与集成融合的OFDM辐射源个体识别方法。首先,从OFDM发射机产生信号的原理出发,对可能产生OFDM发射机指纹差异的来源进行分析,对相邻帧OFDM信号做相干积累,有效提升OFDM信号的信噪比,通过截取OFDM前导信号,减少因传输内容差异所带来的影响,对OFDM前导信号进行功率累加和双谱对角切片信号处理,构建OFDM前导信号的多域数据集;随后,将OFDM前导信号多域数据分别送入具有自动软阈值化去噪和具有跨层连接结构防止梯度消失的DRSN网络进行训练,有效减少噪声对发射机指纹信号的干扰和避免阈值设置不佳所带来识别效果不佳的问题,并且在DRSN网络训练时采用5折交叉验证的策略,防止网络训练中出现过拟合的现象,利用Stacking集成学习思想实现3个DRSN网络初级预测结果的融合;最后,将融合结果作为次级数据送入逻辑回归LR(Logistic Regression)次级线性分类器,利用ECOC(Error Correcting Output Code)策略将多分类任务转为二分类任务,对样本类别进行编码,当测试样本经过二分类器获得一组预测类别编码后,通过计算样本类别编码与预测类别编码之间的欧式距离,根据最小欧式距离所属类别来确定最终分类结果。在公开数据集上的实验结果表明:对比其他深度学习的方法,信噪比为5 dB和0 dB时,DRSN与集成融合的OFDM辐射源识别的准确率分别为97%和95.88%,并且具有较低的复杂度,能够验证在低信噪比下该方法的有效性。

基于灰度均值的自适应FAST角点检测优化算法

光照不均、突变引起的灰度变化会影响图像特征检测效果。为此,设计一种基于灰度均值的自适应FAST-9-12角点检测算法。首先,利用特征点的延展性设计一种小面积双重检测模板,减少像素点与中心点的比较次数,提高区域正检率和检测速度;其次,依据图像局部灰度均值,在每个像素点检测模板内自适应调整阈值,避免灰度变化影响检测效果;最后,根据柔性非极大值抑制的思想设计角点半径抑制原则,筛选鲁棒性更强的角点。在Inria遥感影像数据集上的实验结果表明,FAST-9-12角点检测速度比FAST-12-16,FAST-9-16两种模板提高22%左右,因自适应阈值的提取方式不易受光照影响,检测准确率分别提高4.16和3.11个百分点,实现了图像特征快速和精准检测。

基于增强梯度算子的软阈值宽带频谱感知算法

为了改善信号梯度特征对幅度的损失以及寻求描述信号的最佳尺度问题,提出了一种基于增强梯度算子的软阈值宽带频谱感知算法。通过引入梯度增强算子还原信号幅值特征,结合信号本身梯度特征,使用不同的尺度描述信号梯度增量,得到软阈值判据,进一步加入尺度融合单元,利用硬阈值加软阈值联合判断的方法,得到描述信号的最佳尺度。理论分析和仿真实验结果表明,在高斯信道和瑞利衰落信道下,相较于MPSG算法,所提算法的检测概率和虚警概率均有明显改善,且复杂度更低。通过对比实测数据的检测效果,所提算法更适用于实际工程中。

基于软阈值降噪的脉冲卷积神经网络轴承故障诊断方法

针对工业场景下滚动轴承信号易受噪声干扰,导致故障诊断准确率低和稳定性差的问题,本文提出一种基于软阈值降噪的脉冲卷积神经网络诊断方法。该方法使用软阈值滤波去噪,运用带时间标签的卷积层处理二维信号,增强动态特征提取能力。同时,通过引入IF和LIF神经元实现对时域和频域信息的联合编码,并采用替代梯度法进行端到端训练。实验结果显示,在信噪比为6dB时,所提方法的诊断准确率达100%,在信噪比为-6dB时诊断准确率达77.33%,优于其他常用方法,表明所提方法在噪声下具有良好的诊断效果和稳定性。

预算软约束对地区经济增长的影响及其门槛效应研究

在内外需求日益疲软以及地方债务规模较大的情况下,我国推出了一系列积极的财政措施,这要求政府增加举债规模。本研究从全国范围和东、中、西部地区出发,使用基准回归模型和门槛效应模型对地方政府预算软约束对经济增长的影响进行了实证检验。研究发现,地方政府财政预算软约束显著阻碍了地区经济增长,且二者之间存在显著的门槛效应。异质性分析表明,地方政府预算软约束对经济增长的负向作用会受到区域差异的影响。因此,地方政府应采取优化转移支付制度、促进融资平台的市场化转型、完善过度负债约束机制、转变官员政绩考核理念、因地制宜地制定差异性防范政策等措施稳定地区经济增长。

DRSTN:深度残差软阈值化网络

在采用深度残差等神经网络模型解决图像分类任务时,特征提取过程损失的一些重要特征会影响模型的分类性能。神经网络“端到端”的学习模式带来的黑盒问题,也会限制其在诸多领域的应用和发展。另外,神经网络模型往往需要较长的训练时间。为了提高深度残差网络模型的分类效果和训练效率,引入了模型迁移方法和软阈值化方法,提出了DRSTN(Deep Residual Soft Thresholding Network)网络,并对此网络结构进行微调,生成了不同版本的DRSTN网络。DRSTN网络的性能得益于3个方面的有机整合:1)通过梯度加权类激活映射(Gradients-weighted Class Activation Mapping,Grad-CAM)方法对网络的特征提取进行可视化,根据可视化结果挑选进一步优化的模型;2)基于模型迁移,研究人员不必全新地搭建模型,可以直接在已有的模型上进行优化,能够节省大量训练时间;3)软阈值化作为非线性变换层嵌入到深度残差网络体系结构中,以消除样本中不相关的特征。实验结果表明,在相同训练条件下,DRSTN_KS(3*3)_RB(2:2:2)网络在CIFAR-10数据集上的分类精度相比SKNet-18,ResNet18和ConvNeXt_tiny网络分别提高了15.5%,8.8%和10.9%;该网络也具有一定的泛化性,在MNIST和Fashion MNIST数据集上能够达到快速的迁移效果,分类精度分别达到99.06%和93.15%。

胎—路接触应力传感器的信号去噪方法研究

为了解决应变传感器在测量胎—路接触应力信号时的噪声干扰问题,提出一种基于自适应噪声完备集合经验模态分解(CEEMDAN)和小波软阈值的胎—路接触应力传感器的信号去噪方法。首先,将胎—路接触应力信号CEEMDAN分解为一系列从高频到低频排列的本征模态函数(IMF)分量;再根据能量值法判定出噪声主导IMF分量和信号主导IMF分量,并对噪声主导IMF分量进行小波软阈值去噪;最后,将去噪后的IMF分量和信号主导IMF分量进行重构,得到去噪后的胎—路接触应力信号。将所提方法分别与CEEMDAN去噪方法、小波软阈值去噪方法进行比较,结果表明:所提方法去噪后信号的信噪比(SNR)分别提高了33%和343%,均方根误差(RMSE)分别降低了64%和252%,说明所提方法的去噪效果更佳。

短时傅里叶变换结合DRSN的滚动轴承故障诊断研究

针对滚动轴承在复杂噪声环境下故障分类困难等问题,文章提出一种短时傅里叶变换(STFT)和深度残差收缩网络(DRSN)相结合的轴承故障诊断方法。首先利用短时傅里叶变换对滚动轴承原始振动信号进行时域频域处理得到信息更丰富的故障时频图样本,分为训练集和测试集;将软阈值模块引入到深度残差网络残差块中,其中的残差连接和软阈值模块能够滤除噪声并提取样本特征中的有效信息,输出到分类器上完成端对端的高准确率轴承故障分类。为验证所提方法的可行性,将该方法与其他模型作对比,实验结果表明,该方法在强噪声干扰下能表现出较高的分类性能,稳定性优于其他模型。

基于CWT-IDRSN的风机滚动轴承故障诊断

针对强噪声环境下传统轴承故障诊断方法对故障识别率低,深度残差收缩网络在降噪时对频域信号丢失的问题,提出了一种基于连续小波变换(continuous wavelet transform,CWT)和改进的深度残差收缩网络(improved deep residual shrinkage network,IDRSN)的故障诊断模型。首先,利用CWT将轴承振动信号转换为二维时频图,作为输入样本,用于解决深度残差收缩网络在直接处理振动信号时引起的频域失真问题;其次,设计了一种改进的软阈值函数(improved soft threshold function,ISTF),解决了因软阈值化引起的信号失真,设计了改进的软阈值模块(improved soft threshold block,ISTB)和自适应斜率模块(adaptive slope block,ASB),构建了改进的残差收缩单元(improved residual shrinkage building unit,IRSBU)以实现自适应地确定最佳阈值并进一步调整输出;最后,利用凯斯西储大学滚动轴承数据集与风机轴承振动数据采集实验台收集的滚动轴承数据集对所提方法进行实验验证。结果证明相较于其他方法,所提的故障诊断方法有更好的泛化性和通用性,分类准确率分别达到了99.75%和99.69%。