Signal Processing最新文献_第9页

Minimally informed linear discriminant analysis: Training an LDA model with unlabelled data 最小知情线性判别分析：用未标记数据训练LDA模型

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-07 DOI: 10.1016/j.sigpro.2025.110226

Nicolas Heintz , Tom Francart , Alexander Bertrand

{"title":"Minimally informed linear discriminant analysis: Training an LDA model with unlabelled data","authors":"Nicolas Heintz , Tom Francart , Alexander Bertrand","doi":"10.1016/j.sigpro.2025.110226","DOIUrl":"10.1016/j.sigpro.2025.110226","url":null,"abstract":"<div><div>Linear Discriminant Analysis (LDA) is one of the oldest and most popular linear methods for supervised classification problems. Computing the optimal LDA projection vector requires calculating the average and covariance of the feature vectors of each class individually, which necessitates class labels to estimate these statistics from the data. In this paper we demonstrate that, if some minor prior information is available, it is possible to compute the exact projection vector from LDA models based on unlabelled data. More precisely, we show that either one of the following three pieces of information is sufficient to compute the LDA projection vector if only unlabelled data are available: (1) the class average of one of the two classes, (2) the difference between both class averages (up to a scaling), or (3) the class covariance matrices (up to a scaling). These theoretical results are validated in numerical experiments, demonstrating that this minimally informed Linear Discriminant Analysis (MILDA) model closely approximates the solution of a supervised LDA model, even on high-dimensional, poorly separated or extremely imbalanced data. Furthermore, we show that the MILDA projection vector can be computed in a closed form with a computational cost comparable to LDA and is able to quickly adapt to non-stationary data, making it well-suited to use as an adaptive classifier that is continuously retrained on (unlabelled) streaming data.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110226"},"PeriodicalIF":3.6,"publicationDate":"2025-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144852208","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

Unbiased initial phase estimation for real-valued sinusoids with known frequency via spectral leakage compensation 基于频谱泄漏补偿的已知频率实值正弦波的无偏初始相位估计

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-07 DOI: 10.1016/j.sigpro.2025.110227

Zhe Zhao , Linyue Zhang , Feng Zhang

{"title":"Unbiased initial phase estimation for real-valued sinusoids with known frequency via spectral leakage compensation","authors":"Zhe Zhao , Linyue Zhang , Feng Zhang","doi":"10.1016/j.sigpro.2025.110227","DOIUrl":"10.1016/j.sigpro.2025.110227","url":null,"abstract":"<div><div>This paper investigates the problem of initial phase estimation for a real-valued sinusoidal signal with known frequency. We analyze the bias of the conventional maximum likelihood estimator (MLE) and show that it primarily arises from spectral leakage in the discrete Fourier transform (DFT). Based on this observation, we propose a novel unbiased estimator that eliminates the influence of spectral leakage, thereby achieving unbiased estimation of the initial phase. From a theoretical perspective, we prove that a statistic related to the proposed unbiased estimator is not complete. As a result, it is not possible to theoretically establish that the proposed estimator is the minimum variance unbiased estimator (MVUE) within the framework of the Lehmann–Scheffé theorem, due to the incompleteness of the statistic. Nevertheless, Monte Carlo simulations are conducted to evaluate the performance of the proposed estimator under various frequencies, initial phases, and signal-to-noise ratio (SNR) conditions. The results show that the proposed method consistently achieves unbiased estimation and yields a variance close to the Cramér–Rao lower bound (CRLB) in all tested scenarios.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110227"},"PeriodicalIF":3.6,"publicationDate":"2025-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144813920","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

Error bound for two-dimensional DOA joint estimation in RIS assisted wireless network RIS辅助无线网络中二维DOA联合估计误差界

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-07 DOI: 10.1016/j.sigpro.2025.110229

Cuimin Pan , Xiangbin Yu , Jingjing Pan , Han Zhang

{"title":"Error bound for two-dimensional DOA joint estimation in RIS assisted wireless network","authors":"Cuimin Pan , Xiangbin Yu , Jingjing Pan , Han Zhang","doi":"10.1016/j.sigpro.2025.110229","DOIUrl":"10.1016/j.sigpro.2025.110229","url":null,"abstract":"<div><div>Reconfigurable intelligent surface (RIS) has been a crucial enabler for improving wireless localization accuracy through effectively controlling radio propagation environment. This paper investigates the performance bound for RIS-assisted two-dimensional (2D) direction of arrival (DOA) joint estimation. While the Cramér–Rao lower bound (CRLB) serves as the fundamental performance benchmark for mean square error, it is only asymptotically tight. To this end, an information-theory performance bound termed 2D DOA entropy error (2D-DEE) is proposed through statistical characterization of angle estimation uncertainty. Specifically, the joint <em>a posteriori</em> probability density function (PDF) of 2D DOA is first derived incorporating the uniform and independent <em>a priori</em> distributions of DOAs. Based on this joint <em>a posteriori</em> PDF, the <em>a posteriori</em> entropy is then normalized for different signal-to-noise ratio (SNR) to derive an explicit expression for 2D-DEE. For further insight, the asymptotic expression for entropy errors of 1D DOA and 2D DOA are analyzed in high SNR region. Extensive numerical results validate the accuracy of theoretical analysis and demonstrate that the derived 2D-DEE is able to maintain tight over wider range of SNR in evaluating and predicting 2D DOA estimation performance.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110229"},"PeriodicalIF":3.6,"publicationDate":"2025-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144809427","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

Optimal pilot design for OTFS in linear time-varying channels 线性时变信道OTFS的优化导频设计

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-06 DOI: 10.1016/j.sigpro.2025.110223

Ids van der Werf , Richard Heusdens , Richard C. Hendriks , Geert Leus

引用次数: 0

STWWgram-ODCBAM: Multimodal feature fusion and dynamic attention mechanism for anomalous sound detection STWWgram-ODCBAM：异常声音检测的多模态特征融合与动态注意机制

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-05 DOI: 10.1016/j.sigpro.2025.110218

Libin Zheng, Dongsheng Liu, Tong Wu, Yahui Chen

{"title":"STWWgram-ODCBAM: Multimodal feature fusion and dynamic attention mechanism for anomalous sound detection","authors":"Libin Zheng, Dongsheng Liu, Tong Wu, Yahui Chen","doi":"10.1016/j.sigpro.2025.110218","DOIUrl":"10.1016/j.sigpro.2025.110218","url":null,"abstract":"<div><div>Anomalous sound detection (ASD) aims to identify abnormal acoustic patterns emitted by machines or devices, enabling the timely detection of potential malfunctions. In recent years, various approaches have been proposed to extract both temporal and spectral features from audio data to improve detection performance. However, simply concatenating these features often leads to high-dimensional representations containing redundant information, which increases the risk of overfitting and hinders model performance. To address this issue, we propose a novel model based on a dynamic attention mechanism that adaptively selects and emphasizes informative temporal and spectral features while suppressing irrelevant noise. This enhances the quality of feature representation and improves the accuracy of anomaly detection. Moreover, we design a joint learning architecture that simultaneously captures multimodal features from both time and frequency domains, enabling the model to better capture the complex nature of audio signals and enrich the expressiveness of acoustic features. Experimental results demonstrate that the proposed method significantly outperforms state-of-the-art approaches on the DCASE 2020 Challenge Task 2 dataset, achieving AUC and mAUC improvements of 0.40% and 0.88%, respectively. Notably, for the challenging ToyConveyor machine type, our method achieves a remarkable 5.2% improvement in AUC, demonstrating strong robustness and generalization capability.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110218"},"PeriodicalIF":3.6,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144827086","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

Fast quaternion QR algorithm: Advancing watermarking with multifaceted capabilities 快速四元数QR算法：推进水印与多方面的能力

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-05 DOI: 10.1016/j.sigpro.2025.110215

Yong Chen , Zhigang Jia , Hao Peng , Yaxin Peng , Yan Peng

引用次数: 0

Auto-weighted multi-dimensional feature fusion for incomplete multi-view clustering 不完全多视图聚类的自加权多维特征融合

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-08-04 DOI: 10.1016/j.sigpro.2025.110216

Ao Li , Xinya Xu , Lijuan Zhou , Yanbing Wang , Tianyu Gao

{"title":"Auto-weighted multi-dimensional feature fusion for incomplete multi-view clustering","authors":"Ao Li , Xinya Xu , Lijuan Zhou , Yanbing Wang , Tianyu Gao","doi":"10.1016/j.sigpro.2025.110216","DOIUrl":"10.1016/j.sigpro.2025.110216","url":null,"abstract":"<div><div>Multi-view subspace clustering is an effective method for clustering high-dimensional data but faces several limitations: (1) It often clusters high-dimensional data directly, overlooking the redundancy of original features and the relevance of features across different dimensions. (2) Higher-order correlations and differential structures between views are frequently ignored, leading to suboptimal performance of the fused subspace representation matrix. To address these issues, we propose an auto-weighted multi-dimensional feature fusion incomplete multi-view clustering method (AWMDF<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span>). AWMDF<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> enhances data representation by decomposing the completed data kernel matrix into feature matrices of various dimensions, which are then automatically weighted according to their contribution. These weighted matrices are fused into a consensus feature matrix, which replaces the original high-dimensional data for subspace learning. Additionally, we develop a multi-view subspace fusion method based on the weighted tensor Schatten-p norm, which captures higher-order relationships between views and assigns appropriate weights to each view. AWMDF<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> integrates multi-dimensional feature fusion, subspace learning, and higher-order relational learning into a unified optimization framework. Extensive experiments on six public datasets demonstrate that AWMDF<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> outperforms ten existing advanced baseline methods.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"238 ","pages":"Article 110216"},"PeriodicalIF":3.6,"publicationDate":"2025-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772475","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

Spectrally compatible MIMO radar waveform design for extended target detection 频谱兼容MIMO雷达波形设计扩展目标检测

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-07-31 DOI: 10.1016/j.sigpro.2025.110212

Rongchang Liang , Jinfeng Hu , Yiran Zhang , Dongxu An , Kai Zhong , Jun Liu , Yuankai Wang , Huiyong Li

{"title":"Spectrally compatible MIMO radar waveform design for extended target detection","authors":"Rongchang Liang , Jinfeng Hu , Yiran Zhang , Dongxu An , Kai Zhong , Jun Liu , Yuankai Wang , Huiyong Li","doi":"10.1016/j.sigpro.2025.110212","DOIUrl":"10.1016/j.sigpro.2025.110212","url":null,"abstract":"<div><div>Spectrally compatible MIMO radar waveform design is crucial for extended target detection. Existing methods either maximize the worst-case SINR within target aspect angles (TAAs) while ignoring spectral compatibility, or consider spectral compatibility by maximizing the average SINR, but the worst-case SINR performance may not be ensured. Different from above, under constant modulus (CM) and inequality spectral constraints, we maximize the worst-case SINR within the target aspect angles(TAAs) to satisfy the robustness requirement, while the spectral constraints ensure spectral compatibility by limiting the energy spectral density(ESD) below the predefined threshold. This problem is challenging due to the non-smoothness of max–min design and multiple inequality constraints. Noting that exact penalty terms are suitable for solving inequality constraints, and CM constraints satisfying the complex-circle manifold, we propose the Max-Min-Exact Penalized Product Manifold (MM-E<span><math><msup><mrow><mtext>P</mtext></mrow><mrow><mn>2</mn></mrow></msup></math></span>M) method. First, auxiliary variables are used to address the non-smoothness of max–min design, transforming it into a minimization problem. Next, exact penalty terms are constructed to remove inequality constraints, and the problem is then projected onto the Product-Complex-Circle-Euclidean Manifold (P<span><math><msup><mrow><mtext>C</mtext></mrow><mrow><mn>2</mn></mrow></msup></math></span>EM), thus transforming it into an unconstrained problem. Finally, the parallel simplified quasi-Newton (PSQN) method is divided. The dataset obtained by illuminating a T-72 tank is used to validate the performance. Simulation results demonstrate that the proposed method has following advantages: (i) improves SINR by at least 4.49 dB while controlling energy spectral density exactly; (ii) provides SINR performance that meets requirements for each angle within TAAs.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"238 ","pages":"Article 110212"},"PeriodicalIF":3.6,"publicationDate":"2025-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145117999","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

SKFormer: Diagram captioning via self-knowledge enhanced multi-modal transformer SKFormer：通过自我认知增强的多模态变压器进行图表字幕

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-07-28 DOI: 10.1016/j.sigpro.2025.110208

Xin Hu , Jiaxin Wang , Tao Gao

{"title":"SKFormer: Diagram captioning via self-knowledge enhanced multi-modal transformer","authors":"Xin Hu , Jiaxin Wang , Tao Gao","doi":"10.1016/j.sigpro.2025.110208","DOIUrl":"10.1016/j.sigpro.2025.110208","url":null,"abstract":"<div><div>Diagram captioning aims to generate sentences with the assistance of key visual objects and relationships. This task is the key basis of several applications like cross-modal retrieval and textbook question answering. Most diagrams consist of simple color blocks and geometric shapes, and the knowledge conveyed is professional and diverse. This not only results in high annotation costs, but also exacerbates the gap between sparse visuals and complex semantics. In this paper, we propose a self-knowledge enhanced multi-modal Transformer denoted as SKFormer, which is based on an encoder–decoder architecture. The encoder includes a perception aggregation graph network PAG, a self-knowledge mining module SKM, and a multi-modal semantic interaction module MMSI. The PAG network takes diagram patches as nodes, and the relationships between nodes as edges, integrating visual perception laws to enhance the visual representation. The SKM utilizes multi-modal LLMs and OCR tools to mine implicit and explicit knowledge in diagrams, while the MMSI is used for the interaction between the visual and semantic contents. The enhanced diagram representation is processed by the decoder to generate sentences, assisting learners in mastering its knowledge. By conducting comprehensive experiments on two datasets, we demonstrate that the SKFormer achieves superior performance over the competitors.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"238 ","pages":"Article 110208"},"PeriodicalIF":3.6,"publicationDate":"2025-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144749625","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

Heavy-tailed filtering with reduced sensitivity to inaccurate noise covariance 对不准确噪声协方差敏感性降低的重尾滤波

IF 3.6 2区工程技术

Signal Processing Pub Date : 2025-07-28 DOI: 10.1016/j.sigpro.2025.110209

Yuanchao Qu , Ruicheng Ma , Zhe Gao

引用次数: 0