Signal ProcessingPub Date : 2025-03-04DOI: 10.1016/j.sigpro.2025.109982
Haiquan Zhao, Haolin Wang, Yi Peng
{"title":"Data-reuse recursive least-squares algorithm with Riemannian manifold constraint","authors":"Haiquan Zhao, Haolin Wang, Yi Peng","doi":"10.1016/j.sigpro.2025.109982","DOIUrl":"10.1016/j.sigpro.2025.109982","url":null,"abstract":"<div><div>Actual signals often contain nonlinear manifold structures, but traditional filtering algorithms assume data are embedded in Euclidean space, which makes them less effective when handling complicated noise and manifold data. To address these challenges, Riemannian geometry constraints to the traditional data-reuse recursive least-squares (DR-RLS) algorithm is proposed in this paper. Therefore, a novel adaptive filtering algorithm combining the DR-RLS algorithm with Riemannian manifolds is proposed. This algorithm constrains the filter update process on the Riemannian manifold through exponential mapping, enabling better adaptation to nonlinear manifold data structures. Additionally, the tracking performance and convergence speed of the algorithm are enhanced by data reuse. The convergence and computational complexity of the proposed algorithm on the Riemannian manifold are also analyzed. Finally, the effectiveness of the proposed algorithm relative to other methods is demonstrated through simulation results.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"234 ","pages":"Article 109982"},"PeriodicalIF":3.4,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143641719","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}
Signal ProcessingPub Date : 2025-03-04DOI: 10.1016/j.sigpro.2025.109960
Yangyang Song, Xiaozhen Xie
{"title":"Triple-layer representation of low rank and group sparsity for hyperspectral image denoising","authors":"Yangyang Song, Xiaozhen Xie","doi":"10.1016/j.sigpro.2025.109960","DOIUrl":"10.1016/j.sigpro.2025.109960","url":null,"abstract":"<div><div>Hyperspectral image (HSI) denoising is an essential step in image processing. In the regularization-based approaches for this step, various kinds of prior information are investigated only in the original or one-layer transform domains of HSIs. To sufficiently explore deeper priors, we propose a novel triple-layer representation of low-rankness and group sparsity (TLLRGS) for HSI denoising. This method encodes the prior knowledge of HSIs with two low-rank layers and a single group-sparse layer. Specifically, the globally low rank in the original domain is measured by Tucker decomposition in the first layer. Then, the low rank in the gradient domain is captured via orthogonal transforms, which can be regarded as the second layer of our TLLRGS model. To describe the shared sparse pattern in the subspaces of gradient domains, we design an <span><math><msub><mrow><mi>l</mi></mrow><mrow><mn>2</mn><mo>,</mo><mi>γ</mi></mrow></msub></math></span>-norm with the parameter <span><math><mi>γ</mi></math></span> in the third layer. Additionally, we introduce <span><math><msub><mrow><mi>l</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-norm regularization for complex noise, especially sparse noise. To solve the TLLRGS model, we adopt an iterative approach based on the augmented Lagrange multiplier method. Finally, extensive experimental results involving complex noise removal demonstrate the superiority of the TLLRGS model over several state-of-the-art denoising methods.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109960"},"PeriodicalIF":3.4,"publicationDate":"2025-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548882","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}
Signal ProcessingPub Date : 2025-03-02DOI: 10.1016/j.sigpro.2025.109986
Peng-Jia Zou, Peng-Lang Shui, Xiang Liang
{"title":"Outlier-robust tri-percentile and truncated maximum likelihood estimators of parameters of weibull radar clutter","authors":"Peng-Jia Zou, Peng-Lang Shui, Xiang Liang","doi":"10.1016/j.sigpro.2025.109986","DOIUrl":"10.1016/j.sigpro.2025.109986","url":null,"abstract":"<div><div>Weibull distributions have gained much concern for the versatility in modelling radar clutter such as sea, ground, and weather clutters. Most existing parameter estimation methods are sensitive to outliers and have degraded accuracy in real clutter environments with outliers. This paper proposes two classes of outlier-robust parameter estimators of Weibull distribution. One is the tri-percentile (TriP) estimator, where the shape parameter is estimated from the ratio of two sample percentiles and the scale parameter is estimated from the third sample percentile. The relative root mean square error (RRMSE) of the shape parameter is proved to be independent of the two parameters. Moreover, the optimal position setup of the percentiles is chosen to minimize estimation errors. The other is the iterative truncated maximum likelihood (TML) estimator, which obtains more accurate robust estimates. It is shown that the RRMSE of the shape parameter is also independent of the two parameters. The ML estimator is a special example of the iterative TML estimator. Finally, experiments with simulated data and measured radar data are made to compare the performance of the TriP and TML estimators with that of the ML estimators and other existing estimators in the presence of outliers in data.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"234 ","pages":"Article 109986"},"PeriodicalIF":3.4,"publicationDate":"2025-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143551372","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}
Signal ProcessingPub Date : 2025-03-01DOI: 10.1016/j.sigpro.2025.109959
Mirosław Krzyśko , Łukasz Smaga , Jędrzej Wydra
{"title":"Distance of mean embedding for testing independence of functional data","authors":"Mirosław Krzyśko , Łukasz Smaga , Jędrzej Wydra","doi":"10.1016/j.sigpro.2025.109959","DOIUrl":"10.1016/j.sigpro.2025.109959","url":null,"abstract":"<div><div>We investigate independence testing for functional data, which may be either univariate or multivariate. Broadly speaking, our approach involves first reducing the dimensionality of the functional data using basis expansion and then applying the distance of mean embedding - a flexible measure of independence. We enhance this method for pairwise independence by incorporating marginal aggregation, as well as asymmetric and symmetric aggregation measures, to improve test performance and adapt it to mutual independence testing. Our methods are compared with tests based on distance covariance and the Hilbert–Schmidt independence criterion. To evaluate their effectiveness, we present simulation studies and two real data examples using air pollution and chemometric data sets. The new testing procedures demonstrate favorable finite-sample properties, effectively controlling the type I error rate and exhibiting competitive power, making them viable alternatives to covariance-based tests.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109959"},"PeriodicalIF":3.4,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549002","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}
Signal ProcessingPub Date : 2025-02-27DOI: 10.1016/j.sigpro.2025.109984
Hui Liu , Zhiguo Zhou
{"title":"Deep evidential reasoning rule learning","authors":"Hui Liu , Zhiguo Zhou","doi":"10.1016/j.sigpro.2025.109984","DOIUrl":"10.1016/j.sigpro.2025.109984","url":null,"abstract":"<div><div>Deep learning has achieved great success in the past years. However, due to the uncertainty in the real world, the concerns on building reliable models have been raised. However, most current strategies can't achieve this goal in a unified way. Since the recently developed evidential reasoning rule (ER<sup>2</sup>) which is a general and interpretable probabilistic inference engine can integrate reliability to realize adaptive evidence combination and overall reliability is introduced to measure the credibility of output, it is an ideal strategy to help deep learning build more reliable model. As such, a new deep evidential reasoning rule learning method (DER<sup>2</sup>) is developed in this study. DER<sup>2</sup> consists of training, adaptation and testing stage. In training stage, deep neural network with multiple fully connected layers is trained. In adaptation stage, reliability is introduced to tune the trained model to obtain the adapted output for a given test sample. In testing stage, not only the predictive output probability is obtained, but also the overall reliability is estimated to measure the credibility of model output so that the decision maker can determine whether the predictive results should be trusted or not. Meanwhile, the model output can be interpreted through the case-based way. The experimental results demonstrated that DER<sup>2</sup> can obtain better performance when introducing adaptation stage and a high-quality credibility measurement can be realized through overall reliability as well.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109984"},"PeriodicalIF":3.4,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Signal ProcessingPub Date : 2025-02-27DOI: 10.1016/j.sigpro.2025.109956
Marcos Nascimento , Candice Müller , Kayol S. Mayer
{"title":"Split-complex feedforward neural network for GFDM joint channel equalization and signal detection","authors":"Marcos Nascimento , Candice Müller , Kayol S. Mayer","doi":"10.1016/j.sigpro.2025.109956","DOIUrl":"10.1016/j.sigpro.2025.109956","url":null,"abstract":"<div><div>This paper presents a novel approach for channel equalization and signal detection in generalized frequency division multiplexing (GFDM) systems, designed for dispersive channels and capable of handling nonlinearities. In digital communications systems, deep learning (DL) techniques have emerged as a promising alternative to traditional adaptive digital signal processing. Although DL is a trending topic and has been applied to areas such as beamforming, channel estimation, equalization, and decoding, there is limited research on the use of complex-valued neural networks (CVNN), particularly in the context of GFDM systems. In this work, we propose a joint channel equalization and signal detection approach for GFDM based on the fully connected CVNN split-complex feedforward neural network (SCFNN). The proposed SCFNN effectively equalizes the dispersive 5G channel while concurrently detects the symbols non-orthogonally multiplexed in frequency, handling both scenarios with and without clipping, all within a single SCFNN. Results are compared with classical equalization and detection algorithms, as well as with the fully connected real-valued neural network (RVNN) approach. The proposed SCFNN solution presents superior symbol error rate (SER) performance while maintaining computational complexity on par with conventional methods.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109956"},"PeriodicalIF":3.4,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549001","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}
Signal ProcessingPub Date : 2025-02-27DOI: 10.1016/j.sigpro.2025.109954
Cheng Cheng , Jean-Yves Tourneret , Sinan Yıldırım
{"title":"A variational Bayesian marginalized particle filter for jump Markov nonlinear systems with unknown measurement noise parameters","authors":"Cheng Cheng , Jean-Yves Tourneret , Sinan Yıldırım","doi":"10.1016/j.sigpro.2025.109954","DOIUrl":"10.1016/j.sigpro.2025.109954","url":null,"abstract":"<div><div>This paper studies a new variational Bayesian marginalized particle filter for estimating the state vector of a jump Markov nonlinear system (JMNLS) with unknown measurement noise parameters. Conjugate priors are assigned to the variables indicating the system mode of the JMNLS and the measurement noise parameters, which are regarded as unknown parameters. According to the marginalized particle filter, the unknown parameters are marginalized from the joint posterior distribution of the state and the unknown parameters of the JMNLS. The posterior distribution of the state is then approximated by using an appropriate particle filter, and the posterior distributions of the system mode and the measurement noise parameters conditionally on each state particle are calculated by using variational Bayesian inference. A simulation study is conducted to compare the proposed method with state-of-the-art approaches in the context of a modified nonlinear benchmark model and radar target tracking.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109954"},"PeriodicalIF":3.4,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143535017","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}
Signal ProcessingPub Date : 2025-02-25DOI: 10.1016/j.sigpro.2025.109967
Zhao Xueqing , Ren Fuquan , Sun Haibo , Zhang Yan , Ma Yue , Qi Qinghong
{"title":"SAR-CDL: SAR image interpretable despeckling through convolutional dictionary learning network","authors":"Zhao Xueqing , Ren Fuquan , Sun Haibo , Zhang Yan , Ma Yue , Qi Qinghong","doi":"10.1016/j.sigpro.2025.109967","DOIUrl":"10.1016/j.sigpro.2025.109967","url":null,"abstract":"<div><div>Deep learning-based approaches have shown advantages in the task of despeckling for SAR images. However, it is still difficult to explain due to the black-box nature of deep learning. Deep unfolding methods provide an interpretable alternative to building deep neural networks, which combines traditional iterative optimization methods with deep neural networks for image recovery tasks. In this paper, we propose an unfolded deep convolutional dictionary learning framework (SAR-CDL) for SAR image despeckling. A new variational model based on convolutional dictionary for removing multiplicative noise is proposed. The alternate direction multiplier method combining deep learning method are used to optimize the variational model, which can parameterize the model by deep learning in an end-to-end learning manner and avoid the large workload of the tuning process. The performance of the proposed SAR-CDL is validated on both simulated and real SAR datasets. The experimental results show that the proposed model outperforms many state-of-the-art methods in terms of quantitative metrics and visual quality, with a stronger ability to recover the fine structure and texture of the SAR images. In addition, the proposed SAR-CDL is robust to the size of the training set and can achieve appropriate results while reducing the training dataset.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109967"},"PeriodicalIF":3.4,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143535018","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}
Signal ProcessingPub Date : 2025-02-24DOI: 10.1016/j.sigpro.2025.109918
Yan Ke , Jia Liu , Yiliang Han
{"title":"Two-stage reversible data hiding in encrypted domain with public key embedding mechanism","authors":"Yan Ke , Jia Liu , Yiliang Han","doi":"10.1016/j.sigpro.2025.109918","DOIUrl":"10.1016/j.sigpro.2025.109918","url":null,"abstract":"<div><div>Reversible data hiding in encrypted domain (RDH-ED) can perform encryption and data embedding to simultaneously fulfill the privacy protection and access control. The key distribution in current RDH-ED primarily follows a symmetric mechanism, resulting in limitations in key management and distribution. Therefore, public-key embedding (PKE) mechanism in RDH-ED is proposed to address the limitations, where embedding permission is open to the public while extracting is under control. Then a two-stage RDH-ED scheme with PKE mechanism is designed based on learning with errors (LWE) for images. The algorithm of the first stage is redundancy recoding in LWE encrypted domain (RR-LWE) for the ciphertext encrypted from any a pixel bit. Public embedding key is specially constructed. <span><math><mi>N</mi></math></span> bits data could be embedded per ciphertext. The algorithm of the second stage is difference expansion in LWE encrypted domain (DE-LWE) for the ciphertext of the entire image following RR-LWE. It transfers the bit operations of DE from spatial domain into LWE encrypted domain. We theoretically deduce the necessary conditions for embedding correctness and security. Experimental results demonstrate the outperformed effects in security and efficiency of the proposed algorithms. RR-LWE achieves an embedding capacity up to 24 bits per pixel (bpp) and DE-LWE further enhances that by approximately 0.5 bpp.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109918"},"PeriodicalIF":3.4,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143487454","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}
{"title":"Multi-focus image fusion based on visual depth and fractional-order differentiation operators embedding convolution norm","authors":"Yongli Xian , Guangxin Zhao , Xuejian Chen , Congzheng Wang","doi":"10.1016/j.sigpro.2025.109955","DOIUrl":"10.1016/j.sigpro.2025.109955","url":null,"abstract":"<div><div>Multi-focus image fusion technology integrates the focused regions of multiple source images to produce a single, all-in-focus image. However, existing methods have drawbacks, including image artifacts, color distortion, and ambiguous boundaries. In this paper, a spatial-domain two-stage fusion approach is proposed to address these challenges. In the first stage, a fractional-order differentiation operator embedding convolution norm is proposed to amplify pixel texture, while a weighted fusion is applied to obtain an initial fusion result. Here, the absolute difference map between initial fusion result and source images is used as the focus information, ensuring the accuracy of initial decision map. During the second stage, the source images and pseudo-depth information are jointly constructed the feature vector of K-nearest neighbors matting (KNNM) algorithm to refine the decision map, aiming to obtain final decision map with smoother boundaries. Experimental results indicate that the proposed method outperforms existing representative algorithms in both qualitative and quantitative evaluations.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109955"},"PeriodicalIF":3.4,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143487297","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}