Signal Processing最新文献

{"title":"LiDAR point clouds segmentation in adverse weather conditions","authors":"Yi An,&nbsp;Fan Li","doi":"10.1016/j.sigpro.2025.110256","DOIUrl":"10.1016/j.sigpro.2025.110256","url":null,"abstract":"<div><div>With the rapid advancement of 3D sensor technologies and the widespread use of Light Detection and Ranging (LiDAR) systems, point cloud data has become essential for describing real-world scenes. It is widely applied in various fields, including autonomous driving. Autonomous driving requires precise environmental perception, with point cloud segmentation being one of the core technologies. However, autonomous vehicles face various environmental conditions, such as rain, snow, and fog. These adverse weather conditions introduce disturbances into LiDAR data, resulting in significant challenges for point cloud segmentation. Existing point cloud segmentation algorithms often perform poorly under such conditions. To address this challenge, we propose an adverse weather segmentation network. In our network, the multiscale perception generation module includes components for multiscale feature extraction and multi-spatial pillar feature extraction, aimed at capturing multiscale spatial perceptual features and relationships. Additionally, a multiscale spatial fusion module integrates these features into the encoding stream, effectively enhancing feature representations. During data processing, we add instances to enrich the point cloud with coherent elements, ensuring environmental consistency. Experimental results show that our method achieves 64.1% mIoU under adverse weather conditions, demonstrating better performance than existing state-of-the-art approaches. These results highlight the robustness and effectiveness of the proposed method.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110256"},"PeriodicalIF":3.6,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144894883","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":"Multimodal video summarization based on graph contrastive learning with fine-grained graph interaction","authors":"Guangli Wu ,&nbsp;Miaomiao Wang ,&nbsp;Ning Ma","doi":"10.1016/j.sigpro.2025.110250","DOIUrl":"10.1016/j.sigpro.2025.110250","url":null,"abstract":"<div><div>Video summarization aims to extract key information from videos and select concise and representative clips to form a summary. However, unimodal video summarization methods have difficulty capturing the rich semantic in videos, while existing multimodal methods often face interference from redundant frames and noise during modal fusion, resulting in insufficient cross-modal interaction. Therefore, we introduce a multimodal video summarization model based on graph contrastive learning and fine-grained graph interaction. The model first constructs the video and text as graph structures, and uses a spatial–temporal graph network to collaboratively model spatial–temporal dependencies. Second, the node features of the video and text graph are optimized using graph contrastive learning to eliminate redundant frames and noise. In the cross-modal graph matching, the similarity between the video and text graph is modeled in parallel by introducing multiple semantic perspectives to achieve fine-grained cross-modal interaction. In addition, this paper introduces a graph alignment loss to further constrain the consistency of cross-modal semantic alignment. Finally, extensive experiments on two benchmark datasets, TVSum and SumMe, verify the effectiveness of the DCGM model, which outperforms the current state-of-the-art methods in terms of performance.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110250"},"PeriodicalIF":3.6,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890142","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":"Angle estimation based on coarray tensor completion for bistatic MIMO radar with sparse array","authors":"Wenshuai Wang ,&nbsp;Xianpeng Wang ,&nbsp;Dandan Meng ,&nbsp;Yuehao Guo ,&nbsp;Guan Gui","doi":"10.1016/j.sigpro.2025.110248","DOIUrl":"10.1016/j.sigpro.2025.110248","url":null,"abstract":"<div><div>Currently, the parameter estimation methods for sparse array bistatic multiple-input multiple-output (MIMO) radar utilizing coarray tensors primarily focus on continuous virtual arrays, overlooking the overall potential of the entire virtual coarray. To address this limitation, a parameter estimation method based on coarray tensor completion is proposed for bistatic MIMO radar with sparse arrays. First, a coarray tensor with missing elements is constructed using the virtual difference coarray based on cross-correlation. However, this coarray tensor contains whole slices of missing elements, making it difficult to directly perform tensor completion. Therefore, the coarray tensor is reconstructed to ensure it contains no missing slices. Additionally, to perform tensor completion more effectively, the reconstructed tensor needs to maximize the dispersion-to-percentage ratio (DPR) of the missing elements. Subsequently, the tensor nuclear norm minimization problem is solved to complete the reconstructed tensor. Finally, parallel factor (PARAFAC) decomposition is applied to the completed tensor to obtain the factor matrices, which are then used to estimate the direction of departure (DOD) and direction of arrival (DOA). The proposed algorithm leverages all coarray elements, resulting in improved estimation accuracy. Simulation experiments confirm the superiority of the proposed method.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110248"},"PeriodicalIF":3.6,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144907093","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":"Direction-of-arrival estimation of coherent sources with leaky-wave antennas using spatially filtered interpolation","authors":"Rida Maydani ,&nbsp;Yide Wang ,&nbsp;Julien Sarrazin","doi":"10.1016/j.sigpro.2025.110245","DOIUrl":"10.1016/j.sigpro.2025.110245","url":null,"abstract":"<div><div>With their frequency-beam scanning behavior, leaky-wave antennas (LWAs) are promising solutions to develop accurate and cost-effective direction-of-arrival (DoA) estimation systems. However, DoA estimators such as MUSIC face challenges with coherent sources due to the non-Vandermonde LWA steering matrix. Leveraging the unique radiation properties of LWAs, this paper first divides the entire field of view into several angular sectors, and then introduces a robust and accurate sectorized spatially-filtered interpolation (SFI) method to transform the LWA steering matrix into a Vandermonde matrix in each sector while minimizing the issue of out-of-sector interference. The proposed method thus allows the estimation of DoAs of coherent sources with LWAs. The simulation results show that the DoAs of multiple coherent sources across the entire field-of-view, regardless of their angular sector, can be correctly estimated. The performance of the proposed method is shown to be close to the Cramér–Rao Bound.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110245"},"PeriodicalIF":3.6,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890143","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":"A multi-stage few-shot framework for extensible radar-based human activity recognition","authors":"Keyu Pan,&nbsp;Wei-Ping Zhu,&nbsp;Bo Shi","doi":"10.1016/j.sigpro.2025.110244","DOIUrl":"10.1016/j.sigpro.2025.110244","url":null,"abstract":"<div><div>This paper proposes a novel framework for radar-based indoor human activity recognition (HAR) using a multi-stage few-shot learning (FSL) paradigm. The core of our approach lies in the design of a dynamic feature extraction architecture that exploits wavelet convolution along with depthwise separable convolutions to effectively capture multi-scale and multi-frequency information from radar signals. We also propose a meta-learning-inspired mechanism that dynamically adjusts class weights for unseen categories, thereby enhancing adaptability and recognition accuracy in few-shot scenarios. Extensive experiments on five benchmark datasets demonstrate consistent performance gains over state-of-the-art methods, with substantial improvements observed for both seen and unseen classes. These findings highlight the robustness, scalability, and generalization capability of our framework, underscoring its potential to advance radar-based HAR in complex and diverse environments.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110244"},"PeriodicalIF":3.6,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144865060","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":"Adaptive CFAR detectors for subspace signal in uncertain rank-one interference and Gaussian noise","authors":"Zuozhen Wang ,&nbsp;Peng Wang ,&nbsp;Peng Hao ,&nbsp;Ce Shen ,&nbsp;Fei You","doi":"10.1016/j.sigpro.2025.110247","DOIUrl":"10.1016/j.sigpro.2025.110247","url":null,"abstract":"<div><div>This paper addresses the problem of detecting a subspace signal in the presence of rank-one interference and Gaussian noise. The interference steering vector is uncertain but confined to a known subspace of the observables, with its exact coordinate being unknown. Additionally, the interference subspace is linearly independent of the signal subspace. Given the unknown noise covariance matrix, we assume the availability of noise-only (training) data for estimation purposes. The covariance matrices of the test and training data are either identical, indicating a homogeneous environment (HE), or share a common structure with an unknown scaling factor, suggesting a partially HE (PHE). At the design stage, we employ both one-step and two-step generalized likelihood ratio tests (GLRTs) to derive two detectors for HE and one detector for PHE. These new detectors maintain the constant false alarm rate (CFAR) property. Furthermore, they are compared with existing detectors in terms of computational complexity and detection performance. Specifically, the computational complexity of the new detectors proposed in this paper is comparable to that of existing detectors. Extensive numerical experiments confirm that the new detectors consistently outperform existing ones in terms of detection performance.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110247"},"PeriodicalIF":3.6,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144865059","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":"Mainlobe compound jamming suppression method for airborne radar based on superimposed stepped frequency waveform","authors":"Ming Hou,&nbsp;Wenchong Xie,&nbsp;Wei Chen,&nbsp;Yuanyi Xiong","doi":"10.1016/j.sigpro.2025.110255","DOIUrl":"10.1016/j.sigpro.2025.110255","url":null,"abstract":"<div><div>With the continuous development of modern electronic warfare technology, the main electromagnetic threat to radar has changed from single intentional jamming to compound jamming. The emergence of various new types of compound jamming poses a significant threat to the survival and operation of radars. In particular, modern high-power jammers have already been capable of generating both suppressive jamming and deceptive jamming simultaneously. In order to mitigate the threat of the mainlobe suppression-deception compound jamming and to ensure the normal operation of airborne radar, a mainlobe compound jamming suppression method for airborne radar based on superimposed stepped frequency waveform is proposed. In the first stage, the superimposed stepped frequency waveform is used as the transmitted waveform to obtain the echo data. Then, the frequency points covered by the suppressive jamming are eliminated from the echo data in combination with the prior information. In the second stage, the carrier frequency domain compensation is conducted on the data from which the suppressive jamming has been eliminated, enabling the false targets to be shifted to the same true range bin. The true target and clutter are blocked by the blocking matrix, and then adaptive filtering processing is carried out in the carrier frequency domain to achieve the suppression of deceptive jamming. Finally, the remaining clutter is suppressed by the space-time adaptive processing technology. The simulation results demonstrate that this method exhibits excellent performance in mainlobe compound jamming and clutter suppression.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110255"},"PeriodicalIF":3.6,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925944","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":"Bayesian Geometric-based Interactions Learning Model for Self-aware Autonomous Agents","authors":"Hafsa Iqbal ,&nbsp;Pablo Marin ,&nbsp;Lucio Marcenaro ,&nbsp;David Martin Gomez ,&nbsp;Carlo Regazzioni","doi":"10.1016/j.sigpro.2025.110237","DOIUrl":"10.1016/j.sigpro.2025.110237","url":null,"abstract":"<div><div>Autonomous agents can perceive and act in the environment only by interacting with neighboring agents. This paper proposes a novel probabilistic interaction model that allows the agent to learn the dynamics of the interactive experiences incrementally. The learned model represents the dynamic interaction variables within a joint Generalized State (GS) space. This allows to represent the contextual information of the agent’s experiences in the form of time-varying probabilistic graphical models whose variables are geometrically interpreted as <em>attractors</em> . Such a generative model is defined as Interactive Hierarchical Generalized Dynamic Bayesian Network (IH-GDBN). Various interaction models describing the agents’ experiences are collectively stored as bio-inspired memory layers of the agent’s Autobiographical Memory (AM). Knowledge stored inside AM is accessed by the Bayesian inference method called Interactive Geometrical Markov Jump Particle Filter (IG-MJPF) and is able to make inferences of future interaction states based on learned IH-GDBNs. Moreover, such filters are enriched to detect anomalies as effects of unknown geometrical forces, i.e., deviating from the predictions of a model called as Generalized Errors (GEs). The estimation of GEs allows the agent to learn the new models incrementally by evolving the respective layers of AM to adapt the changing interaction situations. The proposed method is tested in real-time, complex overtaking experiments involving two Autonomous Vehicles (AVs). Future work will extend these experiments to scenarios with more than two vehicles to better reflect multi-agent traffic dynamics. Two different sensory modalities are employed to show, how the AM memory layers can be learned from the exteroceptive, i.e., positional trajectories (called odometry module) and proprioceptive sensors, i.e. steering angle and rotors’ velocity. Performance of the proposed method highlights the detection capabilities as well as the ability to learn explainable incremental successive models within the AM. Codes related to this work can be accessed via <span><span>https://github.com/Hafsa-Iqbal/Interaction-Modeling</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110237"},"PeriodicalIF":3.6,"publicationDate":"2025-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144886516","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":"A distributed frequency-constrained multichannel active noise control algorithm based on an extended penalty factor via circular convolution","authors":"Fei Xu,&nbsp;Ning Han,&nbsp;Tian Zhang","doi":"10.1016/j.sigpro.2025.110240","DOIUrl":"10.1016/j.sigpro.2025.110240","url":null,"abstract":"<div><div>Multichannel active noise control (ANC) systems have demonstrated prominent advantages in low-frequency noise suppression. Although frequency-constrained strategies can improve the adaptability of the system in complex acoustic environments and allow noise reduction while preserving useful ambient sounds, existing implementations remain largely confined to single-channel configurations. In practical multichannel ANC systems, the application of frequency constraints is restricted by the inherent time delay of the frequency transformation. To address this issue, this paper proposes a frequency-constrained algorithm specifically designed for distributed multichannel ANC systems. The proposed algorithm integrates a node-adaptive mechanism with a neighborhood-based information fusion strategy, while introducing an extended penalty factor via circular convolution to achieve frequency constraints without requiring transformations of the frequency domain. Furthermore, to reduce computational complexity, a distributed coordinate descent optimization method is used, improving the practical feasibility of the algorithm in real-world multichannel ANC applications. Finally, the performance of the proposed algorithm is compared with existing algorithms through numerical simulations conducted in various noise environments. The results demonstrate that the proposed algorithm effectively enforces frequency constraints in distributed multichannel ANC systems, ensuring the preservation of target frequency components while effectively suppressing unwanted noise across other frequency bands.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110240"},"PeriodicalIF":3.6,"publicationDate":"2025-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144861039","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":"Three-dimensional sparse random mode decomposition: From theory to application","authors":"Chen Luo ,&nbsp;Tao Chen ,&nbsp;Hongye Su ,&nbsp;Luca Mainardi ,&nbsp;Lei Xie","doi":"10.1016/j.sigpro.2025.110239","DOIUrl":"10.1016/j.sigpro.2025.110239","url":null,"abstract":"<div><div>Non-stationary signal decomposition faces significant challenges when handling modes with crossover instantaneous frequencies. While sparse random mode decomposition (SRMD) offers a novel approach through stochastic time–frequency representations, its two-dimensional framework struggles to disentangle overlapping frequency components. Conversely, the chirplet transform (CT) introduces a three-dimensional time–frequency-chirp rate (TFCR) space to separate such components but suffers from reconstruction inaccuracies due to blurring effects. To address these limitations, this paper proposes a three-dimensional sparse random mode decomposition (3D-SRMD) method that combines SRMD with CT technique. In 3D-SRMD, the random features are lifted from a two-dimensional plane to a three-dimensional (3D) space by introducing one extra chirp rate axis. This enhancement provides an intuitive means of disentangling the frequency components overlapped in the low dimension. A novel random feature generation strategy is further designed to improve approximation accuracy and enhance mode separation capability by combining the 3D ridge detection method. Theoretical analysis reveals the separability of crossover components and derives an approximation bound for the proposed 3D sparse random feature model. Numerical experiments demonstrate the method’s superiority over state-of-the-art techniques in decomposing nonlinear and crossover frequency-modulated modes. This work bridges the gap between theoretical interpretability and practical effectiveness in handling complex multi-component signals.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"239 ","pages":"Article 110239"},"PeriodicalIF":3.6,"publicationDate":"2025-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144894885","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}