Signal Processing最新文献_第5页

Analyses of the tail-ℓ2 minimization for fast and enhanced sparse selections 快速和增强稀疏选择的尾部ℓ2 最小化分析

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-10-04 DOI: 10.1016/j.sigpro.2024.109728

Menglin Ye , Shidong Li , Cheng Cheng , Jun Xian

We investigate the effectiveness and efficiency of the iterative tail-

ℓ_{2}

minimization (tail-

ℓ_{2}

-min) technique for its sparse selection capabilities. We conduct profile analyses on the tail-

ℓ_{2}

-min, establishing the equivalence of the tail-

ℓ_{2}

-min problem to a two-stage profile

ℓ_{2}

formulation, both featuring analytical solutions. The tail null space property (NSP) of sensing matrix

A

is shown to be equivalent to the NSP of the newly defined profile matrix

\tilde{A}

. Besides the error bound analysis for the tail-

ℓ_{2}

-min under the typical tail-NSP condition, a novel error bound of the tail-

ℓ_{2}

-min formulation is also established without relying on NSP or restricted isometry property (RIP) assumptions. It merely contains tractable coefficients of

A

, and offers insights into successful recovery, with the observation of the convergent iterative procedure. Numerical studies and the applications to image reconstruction demonstrate the superiority and fast convergence of the tail-

ℓ_{2}

sparse solution over state-of-the-art sparse selection methodologies. The sparsity level of a signal that the tail-

ℓ_{2}

profile algorithm guarantees the recovery is around 41% higher than that of the basis pursuit algorithm. The analytical solutions of the tail-

ℓ_{2}

method at each iteration also ensure that the tail-

ℓ_{2}

sparse recovery process is notably fast, especially for high dimensions and high sparsity levels.

我们研究了迭代尾ℓ2 最小化（tail-ℓ2-min）技术在稀疏选择能力方面的有效性和效率。我们对尾-ℓ2-min 进行了剖面分析，建立了尾-ℓ2-min 问题与两阶段剖面 ℓ2 表述的等价性，两者都具有分析解。研究表明，传感矩阵 A 的尾部无效空间特性（NSP）等同于新定义的剖面矩阵 Ã 的 NSP。除了在典型的尾部 NSP 条件下对尾部-ℓ2-min 进行误差约束分析外，还建立了尾部-ℓ2-min 公式的新误差约束，而无需依赖 NSP 或受限等距特性（RIP）假设。它仅仅包含了可处理的 A 系数，并通过观察收敛迭代过程提供了成功恢复的见解。数值研究和在图像重建中的应用证明了尾ℓ2 稀疏解比最先进的稀疏选择方法更优越、收敛更快。尾-ℓ2剖面算法所能保证恢复的信号稀疏程度比基追求算法高出约41%。尾ℓ2 方法每次迭代的分析解也确保了尾ℓ2 稀疏恢复过程的速度，尤其是在高维和高稀疏度情况下。

{"title":"Analyses of the tail-ℓ2 minimization for fast and enhanced sparse selections","authors":"Menglin Ye , Shidong Li , Cheng Cheng , Jun Xian","doi":"10.1016/j.sigpro.2024.109728","DOIUrl":"10.1016/j.sigpro.2024.109728","url":null,"abstract":"<div><div>We investigate the effectiveness and efficiency of the iterative tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> minimization (tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>-min) technique for its sparse selection capabilities. We conduct profile analyses on the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>-min, establishing the equivalence of the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>-min problem to a two-stage profile <span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> formulation, both featuring analytical solutions. The tail null space property (NSP) of sensing matrix <span><math><mi>A</mi></math></span> is shown to be equivalent to the NSP of the newly defined profile matrix <span><math><mover><mrow><mi>A</mi></mrow><mrow><mo>̃</mo></mrow></mover></math></span>. Besides the error bound analysis for the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>-min under the typical tail-NSP condition, a novel error bound of the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span>-min formulation is also established without relying on NSP or restricted isometry property (RIP) assumptions. It merely contains tractable coefficients of <span><math><mi>A</mi></math></span>, and offers insights into successful recovery, with the observation of the convergent iterative procedure. Numerical studies and the applications to image reconstruction demonstrate the superiority and fast convergence of the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> sparse solution over state-of-the-art sparse selection methodologies. The sparsity level of a signal that the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> profile algorithm guarantees the recovery is around 41% higher than that of the basis pursuit algorithm. The analytical solutions of the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> method at each iteration also ensure that the tail-<span><math><msub><mrow><mi>ℓ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> sparse recovery process is notably fast, especially for high dimensions and high sparsity levels.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109728"},"PeriodicalIF":3.4,"publicationDate":"2024-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Low-complexity recursive constrained maximum Versoria criterion adaptive filtering algorithm 低复杂度递归受限最大韦尔索里亚准则自适应滤波算法

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-10-03 DOI: 10.1016/j.sigpro.2024.109726

Ji Zhao , Lvyu Li , Qiang Li , Bo Li , Hongbin Zhang

Linearly-constrained adaptive filtering algorithms have emerged as promising candidates for system estimation. The existing methods such as the constrained least mean square algorithm rely on mean square error based learning, which delivers suboptimal performance under non-Gaussian noise environments. Therefore, the recursive constrained maximum Versoria criterion (RCMVC) algorithm has been derived and is robust against impulsive distortions. Nonetheless, RCMVC suffers from a notable computational overhead stemming from matrix inversion operations. To circumvent this issue, utilizing the weighting method and the dichotomous coordinate descent (DCD) iteration method, this paper derives a low-complexity version of the RCMVC algorithm called DCD-RCMVC, which alleviates the requirement of matrix inversion and enhances the estimation accuracy and robustness against non-Gaussian interference. Furthermore, we also present a comprehensive theoretical analysis of the DCD-RCMVC algorithm, encompassing discussions on its equivalence, convergence properties, and computational complexity. Simulations performed for system identification problems indicate that the DCD-RCMVC algorithm outperforms the existing state-of-art approaches.

线性约束自适应滤波算法已成为系统估算的理想候选算法。现有的方法，如约束最小均方算法，依赖于基于均方误差的学习，在非高斯噪声环境下性能不佳。因此，人们提出了递归受限最大韦尔索里亚准则（RCMVC）算法，该算法对脉冲失真具有鲁棒性。然而，RCMVC 算法在矩阵反演操作中存在显著的计算开销。为了规避这一问题，本文利用加权法和二分坐标下降（DCD）迭代法，推导出了一种低复杂度的 RCMVC 算法，称为 DCD-RCMVC，它减轻了矩阵反演的要求，提高了估计精度和对非高斯干扰的鲁棒性。此外，我们还对 DCD-RCMVC 算法进行了全面的理论分析，包括对其等价性、收敛特性和计算复杂性的讨论。针对系统识别问题进行的仿真表明，DCD-RCMVC 算法优于现有的先进方法。

{"title":"Low-complexity recursive constrained maximum Versoria criterion adaptive filtering algorithm","authors":"Ji Zhao , Lvyu Li , Qiang Li , Bo Li , Hongbin Zhang","doi":"10.1016/j.sigpro.2024.109726","DOIUrl":"10.1016/j.sigpro.2024.109726","url":null,"abstract":"<div><div>Linearly-constrained adaptive filtering algorithms have emerged as promising candidates for system estimation. The existing methods such as the constrained least mean square algorithm rely on mean square error based learning, which delivers suboptimal performance under non-Gaussian noise environments. Therefore, the recursive constrained maximum Versoria criterion (RCMVC) algorithm has been derived and is robust against impulsive distortions. Nonetheless, RCMVC suffers from a notable computational overhead stemming from matrix inversion operations. To circumvent this issue, utilizing the weighting method and the dichotomous coordinate descent (DCD) iteration method, this paper derives a low-complexity version of the RCMVC algorithm called DCD-RCMVC, which alleviates the requirement of matrix inversion and enhances the estimation accuracy and robustness against non-Gaussian interference. Furthermore, we also present a comprehensive theoretical analysis of the DCD-RCMVC algorithm, encompassing discussions on its equivalence, convergence properties, and computational complexity. Simulations performed for system identification problems indicate that the DCD-RCMVC algorithm outperforms the existing state-of-art approaches.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109726"},"PeriodicalIF":3.4,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sound source localization via distance metric learning with regularization 通过正则化距离度量学习进行声源定位

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-30 DOI: 10.1016/j.sigpro.2024.109721

Mingmin Liu , Zhihua Lu , Xiaodong Wang , João Paulo J. da Costa , Tai Fei

Sound source localization (SSL) or simply direction of arrival (DOA) classification is an important ingredient in acoustic applications. Traditional model-based algorithms are susceptible to the effects of noise and reverberation, while data-driven deep learning algorithms maintain strong performance across a variety of acoustic circumstances, but typically require a large amount of labeled data. Nevertheless, the existing datasets for SSL are not sufficiently big and diverse to achieve the full potential of deep learning algorithms. Then, it is an imperative work to develop a non-data-hungry algorithm of SSL using small or medium data volume. To this end, we propose a regularized distance metric learning algorithm, that is, by means of the kernel method, we design a nonlinear feature transformation from two aspects: feature points and feature distributions. It transforms the data into a new feature space that brings features of the same class as close as possible and removes features of different classes as far away as possible, which can significantly improve the output of a DOA classifier that follows. Experimental results show that the proposed algorithm outperforms deep learning algorithms in diverse acoustic conditions.

声源定位（SSL）或简单的到达方向（DOA）分类是声学应用中的重要组成部分。传统的基于模型的算法容易受到噪声和混响的影响，而数据驱动的深度学习算法在各种声学环境下都能保持强劲的性能，但通常需要大量的标记数据。然而，现有的 SSL 数据集还不够大、不够多样化，无法充分发挥深度学习算法的潜力。因此，利用中小数据量开发不占用数据的 SSL 算法势在必行。为此，我们提出了一种正则化距离度量学习算法，即通过核方法，从特征点和特征分布两个方面设计一种非线性特征变换。它将数据转换到一个新的特征空间，使同类特征尽可能接近，不同类特征尽可能远离，从而显著提高后续 DOA 分类器的输出结果。实验结果表明，在不同的声学条件下，所提出的算法优于深度学习算法。

{"title":"Sound source localization via distance metric learning with regularization","authors":"Mingmin Liu , Zhihua Lu , Xiaodong Wang , João Paulo J. da Costa , Tai Fei","doi":"10.1016/j.sigpro.2024.109721","DOIUrl":"10.1016/j.sigpro.2024.109721","url":null,"abstract":"<div><div>Sound source localization (SSL) or simply direction of arrival (DOA) classification is an important ingredient in acoustic applications. Traditional model-based algorithms are susceptible to the effects of noise and reverberation, while data-driven deep learning algorithms maintain strong performance across a variety of acoustic circumstances, but typically require a large amount of labeled data. Nevertheless, the existing datasets for SSL are not sufficiently big and diverse to achieve the full potential of deep learning algorithms. Then, it is an imperative work to develop a non-data-hungry algorithm of SSL using small or medium data volume. To this end, we propose a regularized distance metric learning algorithm, that is, by means of the kernel method, we design a nonlinear feature transformation from two aspects: feature points and feature distributions. It transforms the data into a new feature space that brings features of the same class as close as possible and removes features of different classes as far away as possible, which can significantly improve the output of a DOA classifier that follows. Experimental results show that the proposed algorithm outperforms deep learning algorithms in diverse acoustic conditions.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109721"},"PeriodicalIF":3.4,"publicationDate":"2024-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust fusion filter for networked uncertain descriptor systems with colored noise and cyber-attacks 具有彩色噪声和网络攻击的网络不确定描述符系统的鲁棒融合滤波器

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-30 DOI: 10.1016/j.sigpro.2024.109724

Yexuan Zhang, Chenjian Ran, Shuli Sun

The robust fusion filtering problem of multi-sensor networked uncertain descriptor systems (NUDSs) with colored noise, uncertain noise variances and cyber-attacks is investigated. During data transmission in unreliable communication networks, the data can be maliciously attacked by attackers. In other words, the local filters (LFs) may receive false data or may not receive data because of the cyber-attacks. By adopting the singular value decomposition (SVD) method, the original NUDSs can be converted into two reduced-order subsystems with uncertain correlated fictitious white noises, and the cyber-attacks are transformed into the fictitious noises. Cross-covariance matrices between local filtering errors are derived. The robust LFs are obtained according to the minimax robust estimation principle. Under the linear unbiased minimum variance criterion, three weighted fusion algorithms are applied to fuse the LFs. For all allowable uncertainties of noise variances and cyber-attacks, the minimal upper bounds of covariance matrices of the local and distributed fusion filters are guaranteed. The proof of their robustness is established through the minimax estimation principle and Lyapunov equation method. Finally, the correctness and effectiveness of the proposed algorithms are verified by a circuit system example.

研究了具有彩色噪声、不确定噪声方差和网络攻击的多传感器网络不确定描述符系统（NUDS）的鲁棒性融合滤波问题。在不可靠的通信网络中传输数据时，数据可能会受到攻击者的恶意攻击。换句话说，本地滤波器（LF）可能会接收到错误数据，也可能因为网络攻击而接收不到数据。通过采用奇异值分解（SVD）方法，可以将原始 NUDS 转换为两个具有不确定相关虚构白噪声的降阶子系统，并将网络攻击转换为虚构噪声。得出局部滤波误差之间的交叉协方差矩阵。根据最小稳健估计原理得到稳健 LF。在线性无偏最小方差准则下，应用三种加权融合算法对 LFs 进行融合。在噪声方差和网络攻击的所有允许不确定性条件下，保证了本地和分布式融合滤波器协方差矩阵的最小上界。通过最小估计原理和 Lyapunov 方程方法证明了它们的鲁棒性。最后，通过一个电路系统实例验证了所提算法的正确性和有效性。

{"title":"Robust fusion filter for networked uncertain descriptor systems with colored noise and cyber-attacks","authors":"Yexuan Zhang, Chenjian Ran, Shuli Sun","doi":"10.1016/j.sigpro.2024.109724","DOIUrl":"10.1016/j.sigpro.2024.109724","url":null,"abstract":"<div><div>The robust fusion filtering problem of multi-sensor networked uncertain descriptor systems (NUDSs) with colored noise, uncertain noise variances and cyber-attacks is investigated. During data transmission in unreliable communication networks, the data can be maliciously attacked by attackers. In other words, the local filters (LFs) may receive false data or may not receive data because of the cyber-attacks. By adopting the singular value decomposition (SVD) method, the original NUDSs can be converted into two reduced-order subsystems with uncertain correlated fictitious white noises, and the cyber-attacks are transformed into the fictitious noises. Cross-covariance matrices between local filtering errors are derived. The robust LFs are obtained according to the minimax robust estimation principle. Under the linear unbiased minimum variance criterion, three weighted fusion algorithms are applied to fuse the LFs. For all allowable uncertainties of noise variances and cyber-attacks, the minimal upper bounds of covariance matrices of the local and distributed fusion filters are guaranteed. The proof of their robustness is established through the minimax estimation principle and Lyapunov equation method. Finally, the correctness and effectiveness of the proposed algorithms are verified by a circuit system example.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109724"},"PeriodicalIF":3.4,"publicationDate":"2024-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Image deconvolution using hybrid threshold based on modified L1-clipped penalty in EM framework 在 EM 框架中使用基于修改后 L1 截断惩罚的混合阈值进行图像解卷积

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-29 DOI: 10.1016/j.sigpro.2024.109725

Ravi Pratap Singh , Manoj Kumar Singh

Image deconvolution remains a challenging task due to its inherent ill-posedness. While existing algorithms show strong numerical performance, their complexity often complicates analysis and implementation. This paper introduces a computationally efficient image deconvolution method within the expectation maximization (EM) framework. The proposed algorithm alternates between an E-step leveraging the fast Fourier transform (FFT) and an M-step utilizing the discrete wavelet transform (DWT). In the M-step, we introduce a novel L₁-clipped penalty to compute the maximum a posteriori (MAP) estimate, resulting in a hybrid threshold that combines the strengths of soft and hard thresholding. This hybrid threshold is mathematically derived, overcoming the high variance of hard-thresholding and the high bias of soft-thresholding, thus optimizing the trade-off between variance and bias. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art techniques in terms of improved signal-to-noise ratio (ISNR) and peak signal-to-noise ratio (PSNR), as well as visual quality. Notably, the proposed method shows average PSNR improvements of 3.49 dB, 4.23 dB, and 1.44 dB for uniform blur and 0.76 dB, 3.57 dB, and 0.66 dB for Gaussian blur on the Set12, BSD68, and Set14 datasets, respectively.

图像解卷积由于其固有的问题性，仍然是一项具有挑战性的任务。虽然现有算法显示出很强的数值性能，但其复杂性往往使分析和实现变得复杂。本文在期望最大化（EM）框架内介绍了一种计算高效的图像解卷积方法。所提出的算法在利用快速傅立叶变换（FFT）的 E 步和利用离散小波变换（DWT）的 M 步之间交替进行。在 M 步中，我们引入了一种新颖的 L1 截断惩罚来计算最大后验（MAP）估计值，从而产生了一种混合阈值，它结合了软阈值和硬阈值的优点。这种混合阈值是通过数学计算得出的，克服了硬阈值的高方差和软阈值的高偏差，从而优化了方差和偏差之间的权衡。广泛的实验证明，我们的方法在改善信噪比（ISNR）和峰值信噪比（PSNR）以及视觉质量方面明显优于最先进的技术。值得注意的是，在 Set12、BSD68 和 Set14 数据集上，所提出的方法对均匀模糊的平均 PSNR 分别提高了 3.49 dB、4.23 dB 和 1.44 dB，对高斯模糊的平均 PSNR 分别提高了 0.76 dB、3.57 dB 和 0.66 dB。

{"title":"Image deconvolution using hybrid threshold based on modified L1-clipped penalty in EM framework","authors":"Ravi Pratap Singh , Manoj Kumar Singh","doi":"10.1016/j.sigpro.2024.109725","DOIUrl":"10.1016/j.sigpro.2024.109725","url":null,"abstract":"<div><div>Image deconvolution remains a challenging task due to its inherent ill-posedness. While existing algorithms show strong numerical performance, their complexity often complicates analysis and implementation. This paper introduces a computationally efficient image deconvolution method within the expectation maximization (EM) framework. The proposed algorithm alternates between an E-step leveraging the fast Fourier transform (FFT) and an M-step utilizing the discrete wavelet transform (DWT). In the M-step, we introduce a novel L<sub>1</sub>-clipped penalty to compute the maximum a posteriori (MAP) estimate, resulting in a hybrid threshold that combines the strengths of soft and hard thresholding. This hybrid threshold is mathematically derived, overcoming the high variance of hard-thresholding and the high bias of soft-thresholding, thus optimizing the trade-off between variance and bias. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art techniques in terms of improved signal-to-noise ratio (ISNR) and peak signal-to-noise ratio (PSNR), as well as visual quality. Notably, the proposed method shows average PSNR improvements of 3.49 dB, 4.23 dB, and 1.44 dB for uniform blur and 0.76 dB, 3.57 dB, and 0.66 dB for Gaussian blur on the Set12, BSD68, and Set14 datasets, respectively.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109725"},"PeriodicalIF":3.4,"publicationDate":"2024-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Isolated point clutter suppression method for airborne STAP radar in wind farm environment 风电场环境中机载 STAP 雷达的孤立点杂波抑制方法

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-29 DOI: 10.1016/j.sigpro.2024.109723

Yuanyi Xiong , Wenchong Xie , Wei Chen , Ming Hou , Chengyin Liu , Yongliang Wang

With the large-scale construction of wind farms, wind turbine isolated point clutter has an increasingly serious impact on airborne radar target detection performance. Traditional space-time adaptive processing methods cannot suppress wind turbine clutter (WTC) with spectrum broadening characteristics, which may lead to a decrease in target detection probability and an increase in false alarm rate. In this paper, a WTC suppression method for airborne radar based on micro-Doppler features is proposed, and we construct the feature subspace of wind turbine echo to distinguish wind turbine, target, clutter, and noise. First, the Sobel operator is used to process the radar range-Doppler spectrum, and the range cells of the wind turbines are preliminarily judged. Then the smallest of constant false alarm rate (SOCFAR) method is used to further confirm the range cells where the wind turbines are located. Next, Mahalanobis distance is used to estimate the optimal dictionary atomic parameters of WTC, and the updated dictionary atoms are used to construct an orthogonal projection matrix to suppress WTC. Finally, short-range nonstationary clutter and sidelobe clutter are suppressed by space-time adaptive segment processing. On the one hand, the proposed method realizes the accurate positioning of wind turbines through image edge detection and constant false alarm detection. On the other hand, Mahalanobis distance is used to estimate the atomic parameters of the wind turbine dictionary, which ensures the homogeneity of wind turbine samples after clutter suppression. The simulation and measured data results show that the proposed method can significantly reduce the false alarm rate caused by WTC while ensuring the effective detection of the target.

随着风电场的大规模建设，风电机组孤立点杂波对机载雷达目标探测性能的影响日益严重。传统的时空自适应处理方法无法抑制具有频谱展宽特性的风电杂波（WTC），从而可能导致目标探测概率下降和误报率上升。本文提出了一种基于微多普勒特征的机载雷达 WTC 抑制方法，并构建了风轮机回波的特征子空间，以区分风轮机、目标、杂波和噪声。首先，使用 Sobel 算子处理雷达测距-多普勒频谱，初步判断风力涡轮机的测距单元。然后，使用恒定误报率最小法（SOCFAR）进一步确认风力涡轮机所在的测距单元。接着，利用 Mahalanobis 距离估计永利国际娱乐的最佳字典原子参数，并利用更新后的字典原子构建正交投影矩阵来抑制永利国际娱乐。最后，通过时空自适应分段处理来抑制短程非稳态杂波和侧叶杂波。一方面，所提出的方法通过图像边缘检测和恒定误报检测实现了风机的精确定位。另一方面，利用 Mahalanobis 距离估计风机字典的原子参数，确保了杂波抑制后风机样本的同质性。仿真和实测数据结果表明，所提出的方法能显著降低永利国际娱乐造成的误报率，同时确保目标的有效检测。

{"title":"Isolated point clutter suppression method for airborne STAP radar in wind farm environment","authors":"Yuanyi Xiong , Wenchong Xie , Wei Chen , Ming Hou , Chengyin Liu , Yongliang Wang","doi":"10.1016/j.sigpro.2024.109723","DOIUrl":"10.1016/j.sigpro.2024.109723","url":null,"abstract":"<div><div>With the large-scale construction of wind farms, wind turbine isolated point clutter has an increasingly serious impact on airborne radar target detection performance. Traditional space-time adaptive processing methods cannot suppress wind turbine clutter (WTC) with spectrum broadening characteristics, which may lead to a decrease in target detection probability and an increase in false alarm rate. In this paper, a WTC suppression method for airborne radar based on micro-Doppler features is proposed, and we construct the feature subspace of wind turbine echo to distinguish wind turbine, target, clutter, and noise. First, the Sobel operator is used to process the radar range-Doppler spectrum, and the range cells of the wind turbines are preliminarily judged. Then the smallest of constant false alarm rate (SO<img>CFAR) method is used to further confirm the range cells where the wind turbines are located. Next, Mahalanobis distance is used to estimate the optimal dictionary atomic parameters of WTC, and the updated dictionary atoms are used to construct an orthogonal projection matrix to suppress WTC. Finally, short-range nonstationary clutter and sidelobe clutter are suppressed by space-time adaptive segment processing. On the one hand, the proposed method realizes the accurate positioning of wind turbines through image edge detection and constant false alarm detection. On the other hand, Mahalanobis distance is used to estimate the atomic parameters of the wind turbine dictionary, which ensures the homogeneity of wind turbine samples after clutter suppression. The simulation and measured data results show that the proposed method can significantly reduce the false alarm rate caused by WTC while ensuring the effective detection of the target.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109723"},"PeriodicalIF":3.4,"publicationDate":"2024-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

EdgeStereoSR: A multi-task network with transformers for stereo image super-resolution considering edge prior EdgeStereoSR：带变换器的多任务网络，用于考虑边缘先验的立体图像超分辨率

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-28 DOI: 10.1016/j.sigpro.2024.109719

Anqi Liu, Sumei Li, Yongli Chang, Yonghong Hou

Recently, stereo image super-resolution methods focusing on exploring cross-view information have been widely studied and achieved good performance. However, it is still challenging for them to reconstruct high-quality high-frequency details. In addition, they mainly focus on improving quantitative metrics, neglecting the perceptual quality of reconstructed images. In this paper, to improve the accuracy of high-frequency reconstruction, we propose a multi-task network with Transformers considering edge prior, named EdgeStereoSR, which achieves better stereo image SR under the guidance of edge detection. Basically, edge priors have two contributions. First, we propose a cross-view Transformer (CVT), which utilizes edge priors to guide the correspondence search, thus more accurate cross-view information can be captured. Second, we propose a cross-task Transformer (CTT), which exploits edge priors to guide the high-frequency reconstruction, thus images with more details and sharper edges can be reconstructed. To further improve the visual quality, we propose EdgeStereoSR-G, integrating the generative adversarial network into EdgeStereoSR. Specially, a spatial-view discriminator is designed to learn the stereo image distribution so as to make the reconstructed stereo image more photo-realistic and avoid parallax inconsistency. Extensive experiments show that the proposed methods are superior to other state-of-the-art methods in terms of both quantitative metrics and visual quality.

最近，人们广泛研究了以探索跨视角信息为重点的立体图像超分辨率方法，并取得了良好的效果。然而，它们在重建高质量高频细节方面仍面临挑战。此外，这些方法主要侧重于提高定量指标，忽视了重建图像的感知质量。在本文中，为了提高高频重建的准确性，我们提出了一种带有考虑边缘先验的 Transformers 的多任务网络，命名为 EdgeStereoSR，它能在边缘检测的指导下实现更好的立体图像 SR。基本上，边缘先验有两个贡献。首先，我们提出了跨视角变换器（CVT），利用边缘先验来指导对应搜索，从而捕捉到更准确的跨视角信息。其次，我们提出了跨任务变换器（CTT），利用边缘先验来指导高频重建，从而重建出细节更丰富、边缘更清晰的图像。为了进一步提高视觉质量，我们提出了 EdgeStereoSR-G，将生成对抗网络集成到 EdgeStereoSR 中。我们还特别设计了一个空间视角判别器来学习立体图像的分布，从而使重建的立体图像更加逼真，避免视差不一致。大量实验表明，所提出的方法在定量指标和视觉质量方面都优于其他最先进的方法。

{"title":"EdgeStereoSR: A multi-task network with transformers for stereo image super-resolution considering edge prior","authors":"Anqi Liu, Sumei Li, Yongli Chang, Yonghong Hou","doi":"10.1016/j.sigpro.2024.109719","DOIUrl":"10.1016/j.sigpro.2024.109719","url":null,"abstract":"<div><div>Recently, stereo image super-resolution methods focusing on exploring cross-view information have been widely studied and achieved good performance. However, it is still challenging for them to reconstruct high-quality high-frequency details. In addition, they mainly focus on improving quantitative metrics, neglecting the perceptual quality of reconstructed images. In this paper, to improve the accuracy of high-frequency reconstruction, we propose a multi-task network with Transformers considering edge prior, named EdgeStereoSR, which achieves better stereo image SR under the guidance of edge detection. Basically, edge priors have two contributions. First, we propose a cross-view Transformer (CVT), which utilizes edge priors to guide the correspondence search, thus more accurate cross-view information can be captured. Second, we propose a cross-task Transformer (CTT), which exploits edge priors to guide the high-frequency reconstruction, thus images with more details and sharper edges can be reconstructed. To further improve the visual quality, we propose EdgeStereoSR-G, integrating the generative adversarial network into EdgeStereoSR. Specially, a spatial-view discriminator is designed to learn the stereo image distribution so as to make the reconstructed stereo image more photo-realistic and avoid parallax inconsistency. Extensive experiments show that the proposed methods are superior to other state-of-the-art methods in terms of both quantitative metrics and visual quality.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109719"},"PeriodicalIF":3.4,"publicationDate":"2024-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418350","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adaptive radar target detection in nonzero-mean compound Gaussian sea clutter with random texture 具有随机纹理的非零均值复合高斯海杂波中的自适应雷达目标探测

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-28 DOI: 10.1016/j.sigpro.2024.109720

Haoqi Wu, Zhihang Wang, Hongzhi Guo, Zishu He

This paper deals with the radar target detecting problem in nonzero-mean compound Gaussian sea clutter with random texture. The texture is considered to be an inverse Gamma, Gamma, or inverse Gaussian variable. Three novel adaptive detectors using the two-step maximum a posteriori (MAP) generalized likelihood ratio test (GLRT) are proposed. More precisely, we derive the test statistics of the proposed detectors for known mean vector (MV) and speckle covariance matrix (CM) in the first step. In the second step, unbiased and consistent estimators are proposed to estimate the MV and CM in nonzero-mean compound Gaussian circumstances. We acquire the fully adaptive nonzero-mean GLRT detectors by substituting the estimates into the test statistics. Then, the constant false alarm rate (CFAR) properties of the proposed detectors with respect to (w.r.t.) the speckle CM are proved. Finally, the performance of three proposed detectors is verified by simulation experiments using the synthetic and real sea clutter data.

本文讨论了在具有随机纹理的非零均值复合高斯海杂波中的雷达目标探测问题。纹理被认为是反伽马、伽马或反高斯变量。我们提出了三种使用两步最大后验（MAP）广义似然比检验（GLRT）的新型自适应探测器。更确切地说，我们在第一步推导出了已知均值向量（MV）和斑点协方差矩阵（CM）的检测统计量。第二步，提出无偏且一致的估计器，以估计非零均值复合高斯情况下的 MV 和 CM。通过将估计值代入测试统计量，我们获得了完全自适应的非零均值 GLRT 检测器。然后，证明了所提出的探测器相对于（相对于）斑点 CM 的恒定误报率（CFAR）特性。最后，通过使用合成和真实海杂波数据进行模拟实验，验证了所提出的三种探测器的性能。

引用次数: 0

Protecting copyright of stable diffusion models from ambiguity attacks 保护稳定扩散模型版权免受模糊攻击

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-28 DOI: 10.1016/j.sigpro.2024.109722

Zihan Yuan, Li Li, Zichi Wang, Xinpeng Zhang

In recent years, the stable diffusion models (SDMs) have been widely used in text-to-image generative tasks, and their copyright protection problem has been concerned by scholars. The model owners can embed watermarks into SDMs by fine-tuning them, and use the prompt-watermark pair to complete model ownership authentication. However, the attackers can obfuscate model ownership by forging the relationship between the fake prompt and the watermark image. Therefore, this paper proposes a black-box copyright protection method for SDMs, which can effectively resist watermark ambiguity attacks. Specifically, we adopt an irreversible watermarking technology to complete watermark embedding. The hash function is used to ensure the unidirectional irreversible generation of the trigger prompts using the secret key. Then, the trigger set consisting of trigger prompts and watermarks is used to fine-tune the SDMs to embed the watermarks. Without the secret key, it is not possible for the attackers to reverse build the specific prompts with internal associations. Experiments show that our method can protect the copyright of SDMs effectively and resist ambiguity attacks without the model performance degradation.

近年来，稳定扩散模型（SDM）在文本到图像的生成任务中得到了广泛应用，其版权保护问题也受到了学者们的关注。模型所有者可以通过微调在 SDM 中嵌入水印，并利用提示-水印对完成模型所有权认证。然而，攻击者可以通过伪造假提示和水印图像之间的关系来混淆模型所有权。因此，本文提出了一种 SDM 的黑盒版权保护方法，可以有效抵御水印模糊攻击。具体来说，我们采用不可逆水印技术完成水印嵌入。利用哈希函数确保使用秘钥单向不可逆地生成触发提示。然后，利用由触发提示和水印组成的触发集对 SDM 进行微调，嵌入水印。如果没有秘钥，攻击者就不可能反向生成具有内部关联的特定提示。实验表明，我们的方法可以有效保护 SDMs 的版权，并在不降低模型性能的情况下抵御歧义攻击。

{"title":"Protecting copyright of stable diffusion models from ambiguity attacks","authors":"Zihan Yuan, Li Li, Zichi Wang, Xinpeng Zhang","doi":"10.1016/j.sigpro.2024.109722","DOIUrl":"10.1016/j.sigpro.2024.109722","url":null,"abstract":"<div><div>In recent years, the stable diffusion models (SDMs) have been widely used in text-to-image generative tasks, and their copyright protection problem has been concerned by scholars. The model owners can embed watermarks into SDMs by fine-tuning them, and use the prompt-watermark pair to complete model ownership authentication. However, the attackers can obfuscate model ownership by forging the relationship between the fake prompt and the watermark image. Therefore, this paper proposes a black-box copyright protection method for SDMs, which can effectively resist watermark ambiguity attacks. Specifically, we adopt an irreversible watermarking technology to complete watermark embedding. The hash function is used to ensure the unidirectional irreversible generation of the trigger prompts using the secret key. Then, the trigger set consisting of trigger prompts and watermarks is used to fine-tune the SDMs to embed the watermarks. Without the secret key, it is not possible for the attackers to reverse build the specific prompts with internal associations. Experiments show that our method can protect the copyright of SDMs effectively and resist ambiguity attacks without the model performance degradation.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"227 ","pages":"Article 109722"},"PeriodicalIF":3.4,"publicationDate":"2024-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142418356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-objective network resource allocation method based on fractional PID control 基于分数 PID 控制的多目标网络资源分配方法

IF 3.4 2区工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC

Signal Processing

Pub Date : 2024-09-25 DOI: 10.1016/j.sigpro.2024.109717

Xintong Ni, Yiheng Wei, Shuaiyu Zhou, Meng Tao

In this paper, a fractional proportional–integral–derivative (PID) distributed optimization algorithm is proposed to solve the network resource allocation problem. The algorithm combines fractional calculus and the concept of PID control, which improves the convergence rate and increases the freedom, flexibility and potential with multiple parameters compared with the existing algorithms. Meanwhile, the results of simulation study verified the efficiency and superiority of the algorithm.

本文提出了一种分数比例积分派生（PID）分布式优化算法来解决网络资源分配问题。该算法结合了分数微积分和 PID 控制的概念，与现有算法相比，提高了收敛速度，增加了多参数的自由度、灵活性和潜力。同时，仿真研究结果验证了该算法的高效性和优越性。

引用次数: 0