DeepSVDNet:A Deep Learning-Based Approach for Detecting and Classifying Vision-Threatening Diabetic Retinopathy in Retinal Fundus Images

Artificial Intelligence(AI)is being increasingly used for diagnosing Vision-Threatening Diabetic Retinopathy(VTDR),which is a leading cause of visual impairment and blindness worldwide.However,previous automated VTDR detection methods have mainly relied on manual feature extraction and classification,leading to errors.This paper proposes a novel VTDR detection and classification model that combines different models through majority voting.Our proposed methodology involves preprocessing,data augmentation,feature extraction,and classification stages.We use a hybrid convolutional neural network-singular value decomposition(CNN-SVD)model for feature extraction and selection and an improved SVM-RBF with a Decision Tree(DT)and K-Nearest Neighbor(KNN)for classification.We tested our model on the IDRiD dataset and achieved an accuracy of 98.06%,a sensitivity of 83.67%,and a specificity of 100%for DR detection and evaluation tests,respectively.Our proposed approach outperforms baseline techniques and provides a more robust and accurate method for VTDR detection.

基于QR迭代的量子奇异值分解

针对大型矩阵奇异值分解(singular value decomposition,SVD)时使用经典算法时间复杂度较高,以及已有的量子SVD算法要求待分解的矩阵必须具有非稀疏低秩的性质,并且在计算过程中构造任意大小酉矩阵对目前的量子计算机来说实现起来并不容易等问题,提出基于QR迭代的量子SVD。QR迭代使用的是Householder变换,通过量子矩阵乘法运算完成经典矩阵乘法运算过程。实验结果表明,该方法能够得到所求矩阵的奇异值及奇异矩阵,使大型矩阵的SVD具有可行性。

基于整数小波变换和SVD的视频水印算法

提出了一种以二值图像为水印的混合整数小波变换和奇异值分解的视频水印盲提取算法。对水印图像进行混沌加密和Arnold置乱处理,选择计算复杂度低的直方图算法将视频分割为若干场景;借助密钥随机选取某些场景的亮度分量进行l级整数小波变换,再对低频子带进行分块的奇异值分解;采用量化的方法,将预处理后的水印图像嵌入奇异值分解后的最大奇异值中。在嵌入了水印的视频场景中提取所有的水印版本之后,利用对提取的所有水印信号版本进行统计求和的方法得到最终提取的水印图像。实验表明,提出的算法具有较好的透明性,对常见的处理具有较好的鲁棒性。

基于SVD和TKEO的轴承振动信号特征提取

为了解决滚动轴承振动信号中微弱故障信息难以提取的问题,提出了一种基于奇异值分解(Singular Value Decomposition,SVD)和Teager-Kaiser能量算子(Teager-Kaiser Energy Operator,TKEO)的轴承振动信号特征提取方法。采用SVD将突变信息从背景噪声和光滑信号中分离,提取信号的突变信息;利用TKEO计算突变信息的瞬时能量,对该能量信号进行频谱分析,从而提取出轴承振动信号的能量频谱特征,用于故障检测。将该方法应用于轴承外圈、内圈局部故障状态下的振动信号特征提取,利用特征信息能够准确检测并识别出故障类型,表明了该方法的可行性和有效性。

Contourlet watermarking algorithm based on Arnold scrambling and singular value decomposition

A new digital watermarking algorithm based on the contourlet transform is proposed to improve the robustness and anti-attack performances of digital watermarking. The algorithm uses the Arnold scrambling technique and the singular value decomposition （SVD） scheme. The Arnold scrambling technique is used to preprocess the watermark, and the SVD scheme is used to find the best suitable hiding points. After the contourlet transform of the carrier image, intermediate frequency sub-bands are decomposed to obtain the singularity values. Then the watermark bits scrambled in the Arnold rules are dispersedly embedded into the selected SVD points. Finally, the inverse contourlet transform is applied to obtain the carrier image with the watermark. In the extraction part, the watermark can be extracted by the semi-blind watermark extracting algorithm. Simulation results show that the proposed algorithm has better hiding and robustness performances than the traditional contourlet watermarking algorithm and the contourlet watermarking algorithm with SVD. Meanwhile, it has good robustness performances when the embedded watermark is attacked by Gaussian noise, salt- and-pepper noise, multiplicative noise, image scaling and image cutting attacks, etc. while security is ensured.

高速列车万向轴动不平衡检测的EEMD-Hankel-SVD方法

针对聚合经验模式分解(Ensemble empirical model decomposition,EEMD)的等效滤波特性依然存在模式分量间频带重叠较大的根本缺陷,提出一种高速列车万向轴动不平衡动态检测的新方法。该方法的核心是对万向节安装机座的振动信号进行EEMD分解得到基本模式分量,应用基本模式分量信号来构造Hankel矩阵,对该矩阵进行正交化奇异值(Singular value decomposition,SVD)分解,以奇异值关键叠层作为奇异值的选择准则对信号进行重构,应用重构信号的傅里叶谱来检测高速列车万向轴的动不平衡,消除EEMD分解模式频带重叠对故障特征的淹没和混淆效应,提高了谱的清晰度,凸显了故障特征。应用万向轴动不平衡试验数据对该方法进行试验验证,结果表明,该方法能够有效检测万向轴动不平衡引起的故障特征和万向轴的固有振动特征,与纯EEMD方法相比,该方法在谱的清晰度和故障表征力上得到了显著提高。

The Singular Value Decomposition Analysis between Summer Precipitation in the Dongting Lake Region and the Global Sea Surface Temperature

By dint of the summer precipitation data from 21 stations in the Dongting Lake region during 1960-2008 and the sea surface temperature(SST) data from NOAA,the spatial and temporal distributions of summer precipitation and their correlations with SST are analyzed.The coupling relationship between the anomalous distribution in summer precipitation and the variation of SST has between studied with the Singular Value Decomposition(SVD) analysis.The increase or decrease of summer precipitation in the Dongting Lake region is closely associated with the SST anomalies in three key regions.The variation of SST in the three key regions has been proved to be a significant previous signal to anomaly of summer rainfall in Dongting region.

Dynamic unbalance detection of cardan shaft in high-speed train based on EMD-SVD-NHT

Contrary to the aliasing defect between the adjacent intrinsic model functions(IMFs) existing in empirical model decomposition(EMD), a new method of detecting dynamic unbalance with cardan shaft in high-speed train was proposed by applying the combination between EMD, Hankel matrix, singular value decomposition(SVD) and normalized Hilbert transform(NHT). The vibration signals of gimbal installed base were decomposed through EMD to get different IMFs. The Hankel matrix constructed through the single IMF was orthogonally executed through SVD. The critical singular values were selected to reconstruct vibration signs on the basis of the key stack of singular values. Instantaneous frequencys(IFs) of reconstructed vibration signs were applied to detect dynamic unbalance with shaft and eliminated clutter spectrum caused by the aliasing defect between the adjacent IMFs, which highlighted the failure characteristics. The method was verified by test data in the unbalance condition of dynamic cardan shaft. The results show that the method effectively detects the fault vibration characteristics caused by cardan shaft dynamic unbalance and extracts the nature vibration features. With comparison to the traditional EMD-NHT, clarity and failure characterization force are significantly improved.

A Scheme of Fragile Watermarking Based on SVD and 2D Chaotic Mapping

This paper proposed a novel fragile watermarking scheme based on singular value decomposition (SVD) and 2D chaotic mapping. It obtains chaotic initial values from the image blocks singular value decomposition and the user’s key, then uses the chaotic mapping to get the chaotic sequence and inserts the sequence into the LSBs of the image blocks to get the watermarked image blocks. The paper reconstructed the watermarked image from all the embedded blocks. The analysis and experimental results show that the scheme is pretty fragile to tampering, and it can localize the tampering position accurately, reach 3×3 blocks.

Analysis of heart rate variability based on singular value decomposition entropy

Assessing the dynamics of heart rate fluctuations can provide valuable information about heart status. In this study, regularity of heart rate variability （HRV） of heart failure patients and healthy persons using the concept of singular value decomposition entropy （SvdEn） is analyzed. SvdEn is calculated from the time series using normalized singular values. The advantage of this method is its simplicity and fast computation. It enables analysis of very short and non-stationary data sets. The results show that SvdEn of patients with congestive heart failure （CHF） shows a low value （SvdEn： 0.056±0.006, p 〈 0.01） which can be completely separated from healthy subjects. In addition, differences of SvdEn values between day and night are found for the healthy groups. SvdEn decreases with age. The lower the SvdEn values, the higher the risk of heart disease. Moreover, SvdEn is associated with the energy of heart rhythm. The results show that using SvdEn for discriminating HRV in different physiological states for clinical applications is feasible and simple.

SVD-LSSVM and its application in chemical pattern classification

Pattern classification is an important field in machine learning; least squares support vector machine (LSSVM) is a powerful tool for pattern classification. A new version of LSSVM, SVD-LSSVM, to save time of selecting hyper parameters for LSSVM is proposed. SVD-LSSVM is trained through singular value decomposition (SVD) of kernel matrix. Cross validation time of selecting hyper parameters can be saved because a new hyper parameter, singular value contribution rate (SVCR), replaces the penalty factor of LSSVM. Several UCI benchmarking data and the Olive classification problem were used to test SVD-LSSVM. The result showed that SVD-LSSVM has good performance in classification and saves time for cross validation.

Detection and correction of level echo based on generalized S-transform and singular value decomposition

The echo of the material level is non-stationary and contains many singularities.The echo contains false echoes and noise,which affects the detection of the material level signals,resulting in low accuracy of material level measurement.A new method for detecting and correcting the material level signal is proposed,which is based on the generalized S-transform and singular value decomposition(GST-SVD).In this project,the change of material level is regarded as the low speed moving target.First,the generalized S-transform is performed on the echo signals.During the transformation process,the variation trend of window of the generalized S-transform is adjusted according to the frequency distribution characteristics of the material level echo signal,achieving the purpose of detecting the signal.Secondly,the SVD is used to reconstruct the time-frequency coefficient matrix.At last,the reconstructed time-frequency matrix performs an inverse transform.The experimental results show that the method can accurately detect the material level echo signal,and it can reserve the detailed characteristics of the signal while suppressing the noise,and reduce the false echo interference.Compared with other methods,the material level measurement error does not exceed 4.01%,and the material level measurement accuracy can reach 0.40%F.S.

Hand-eye calibration with a new linear decomposition algorithm

To solve the homogeneous transformation equation of the form AX=XB in hand-eye calibration, where X represents an unknown transformation from the camera to the robot hand, and A and B denote the known movement transformations associated with the robot hand and the camera, respectively, this paper introduces a new linear decomposition algorithm which consists of singular value decomposition followed by the estimation of the optimal rotation matrix and the least squares equation to solve the rotation matrix of X. Without the requirements of traditional methods that A and B be rigid transformations with the same rotation angle, it enables the extension to non-rigid transformations for A and B. The details of our method are given, together with a short discussion of experimental results, showing that more precision and robustness can be achieved.

The Singular Value Decomposition as a Tool of Investigating Central MHD Instabilities in the HL-1M Tokamak

A variety of strong MHD instabilities are always resulted from MHD activity of Tokamak plasmas. Central MHD instabilities can be observed with pinhole cameras to record soft x-ray (SXR) emission from the plasma along many chords with a high temporal resolution. The investigation of MHD instabilities often necessitates an analysis on spatial-temporal signals. The method of Singular Value Decomposition (SVD) can split such signals into orthogonal spatial and temporal vectors. By this means, the repetition time and the characteristic radius of various MHD phenomena such as sawteeth and snake-like perturbation can be obtained. Moreover, the (1,1) MHD mode is analyzed in great detail by SVD and used to determine the radius of the q = 1 surface.

Shafting misalignment fault diagnosis by means of motor speed signal and SVD-HT method

Aiming at the deficiency of diagnosis method based on vibration signal,a novel method based on speed signal with singular value decomposition and Hilbert transform(SVD-HT)is proposed.The fault diagnosis mechanism based on the speed signal is obtained by constructing the shaft misalignment fault model firstly.Then the SVD-HT method is applied to the processing of the speed signal.The accuracy of the SVD-HT method is verified by comparing the diagnosis results of the order spectrum method and the SVD-HT method.After that,the diagnosis results based on vibration signal and speed signal under no-load and load patterns are compared.Under the no-load pattern,the amplitudes of the speed signal components f_(r),2f_(r) and 4f_(r) are linear with the misalignment.In addition,under the load pattern,the amplitudes of the speed signal components f_(r),2f_(r) and 4f_(r) have a linear relationship with the load.However,the diagnosis result of the vibration signal does not have the above characteristics.The comparison results verify the robustness and reliability of the speed signal and SVD-HT method.The method presented in this paper provides a novel way for misalignment fault diagnosis.

Mobility and equilibrium stability analysis of pin-jointed mechanisms with equilibrium matrix SVD

Under certain load pattern, the geometrically indeterminate pin-jointed mechanisms will present certain shapes to keep static equalization. This paper proposes a matrix-based method to determine the mobility and equilibrium stability of mechanisms according to the effects of the external loads. The first and second variations of the potential energy function of mechanisms under conservative force field are analyzed. Based on the singular value decomposition (SVD) method, a new crite- rion for the mobility and equilibrium stability of mechanisms can be concluded by analyzing the equilibrium matrix. The mobility and stability of mechanisms can be classified by unified matrix formulae. A number of examples are given to demonstrate the proposed criterion. In the end, criteria are summarized in a table.

Effects of surface heating on precipitation over the Tibetan Plateau and its eastern margin

The high terrain of the Tibetan Plateau(TP)has a very important impact on the weather and climate of China,East Asia,South Asia,and even the Northern Hemisphere.However,in recent years,the reasons for the decrease in precipitation in the southeastern edge of the plateau have resulted in cutting-edge research regarding the impact of the TP and its surrounding areas on downstream weather and climate.In this study,the spatial and temporal distribution of surface heat flux and precipitation were analyzed from 1998 to 2022,and the possible mechanism of the decrease of precipitation in the eastern edge of the plateau is explored.The main conclusions are as follows:The annual average sensible heat flux in the TP and its east side is positive,with an average of 33.73 W/m^(2).The annual average latent heat flux is positive,with an average of 42.71 W/m^(2).Precipitation has a similar annual average and seasonal distribution,with modest amounts in the northwest and substantial amounts in the southeast.The average annual accumulated precipitation is 670.69 mm.The first mode of the Empirical Orthogonal Function(EOF)shows that sensible heat flux decreases first,then increases,and then finally decreases during 1998–2022.The modes show the opposite trend in middle part of the plateau.The latent heat flux initially decreases,then increases,and finally decreases in the western plateau and near Sichuan Basin.The mode,however,displays the opposite tendency throughout the rest of the region.The precipitation in the north and south sides of the plateau has decreased since 2013,which is consistent with the changing trend of sensible heat flux.In the rest of the region,the change trend is not obvious.The sensible heat of the main body of the plateau and its east side and Sichuan Basin is negatively correlated with precipitation,that is,when sensible heat flux of the main body of the plateau and its east side and Sichuan Basin is more(less),local precipitation is less(more).The latent heat of the main body of the plateau and its east side,Sichuan Basin is positively correlated with precipitation,indicating that when latent heat flux of the main body of the plateau and its east side,Sichuan Basin is more(less),local precipitation is more(less).

Effective and Efficient Video Compression by the Deep Learning Techniques

Deep learning has reached many successes in Video Processing.Video has become a growing important part of our daily digital interactions.The advancement of better resolution content and the large volume offers serious challenges to the goal of receiving,distributing,compressing and revealing highquality video content.In this paper we propose a novel Effective and Efficient video compression by the Deep Learning framework based on the flask,which creatively combines the Deep Learning Techniques on Convolutional Neural Networks(CNN)and Generative Adversarial Networks(GAN).The video compression method involves the layers are divided into different groups for data processing,using CNN to remove the duplicate frames,repeating the single image instead of the duplicate images by recognizing and detecting minute changes using GAN and recorded with Long Short-Term Memory(LSTM).Instead of the complete image,the small changes generated using GAN are substituted,which helps with frame-level compression.Pixel wise comparison is performed using K-nearest Neighbours(KNN)over the frame,clustered with K-means and Singular Value Decomposition(SVD)is applied for every frame in the video for all three colour channels[Red,Green,Blue]to decrease the dimension of the utility matrix[R,G,B]by extracting its latent factors.Video frames are packed with parameters with the aid of a codec and converted to video format and the results are compared with the original video.Repeated experiments on several videos with different sizes,duration,Frames per second(FPS),and quality results demonstrated a significant resampling rate.On normal,the outcome delivered had around a 10%deviation in quality and over half in size when contrasted,and the original video.

A Hybrid Air Quality Prediction Model Based on Empirical Mode Decomposition

Air pollution is a severe environmental problem in urban areas.Accurate air quality prediction can help governments and individuals make proper decisions to cope with potential air pollution.As a classic time series forecasting model,the AutoRegressive Integrated Moving Average(ARIMA)has been widely adopted in air quality prediction.However,because of the volatility of air quality and the lack of additional context information,i.e.,the spatial relationships among monitor stations,traditional ARIMA models suffer from unstable prediction performance.Though some deep networks can achieve higher accuracy,a mass of training data,heavy computing,and time cost are required.In this paper,we propose a hybrid model to simultaneously predict seven air pollution indicators from multiple monitoring stations.The proposed model consists of three components:(1)an extended ARIMA to predict matrix series of multiple air quality indicators from several adjacent monitoring stations;(2)the Empirical Mode Decomposition(EMD)to decompose the air quality time series data into multiple smooth sub-series;and(3)the truncated Singular Value Decomposition(SvD)to compress and denoise the expanded matrix.Experimental results on the public dataset show that our proposed model outperforms the state-of-art air quality forecasting models in both accuracy and time cost.

伪谱计算的增广块Householder Arnoldi算法（英文）

伪谱是解释非正规矩阵或算子行为的一个有用工具.矩阵伪谱计算的一个常用方法是grid-SVD算法,实现这个算法需要在每一个网格点处作奇异值分解(SVD);另外一个计算方法是基于Schur分解的逆Lanczos算法.由于上述方法的计算量比较大,通常只适用于中小型矩阵.近些年,有些学者探讨了大规模矩阵伪谱计算的Krylov子空间投影方法.在探讨了Householder Arnoldi(HA)算法块情形的计算行为和实用性能的基础上,提出了计算大规模矩阵伪谱的增广块HA(ABHA)算法,并对一些典型测试矩阵进行了一系列的数值试验.数值结果表明,增广块HA(ABHA)算法比HA算法,块隐式重启Arnoldi(BLIRA)算法和逆Lanczos算法的计算效率更高,更具优越性.