Digital Signal Processing最新文献

{"title":"Compressive sensing networks based on attention mechanism reconfiguration","authors":"Yuhui Gao ,&nbsp;Jingyi Liu ,&nbsp;Hao Peng ,&nbsp;Shiqiang Chen","doi":"10.1016/j.dsp.2025.105413","DOIUrl":"10.1016/j.dsp.2025.105413","url":null,"abstract":"<div><div>The combination of deep learning and compressive sensing has brought new breakthroughs in the field of image and video processing, but how to design compressive sensing networks with good generalization ability and low computational complexity is still a great challenge. In this paper, we propose a multiscale compressive sensing network reconstructed based on the attention mechanism, where training a single model allows sampling and reconstruction of arbitrary sampling ratios. Initially, in the sampling phase, we employ multi-scale adaptive sampling within the wavelet domain. This method dynamically adjusts the sampling ratios of various image blocks to accommodate the varying complexities of different regions through a multi-scale mechanism, thereby enhancing data utilization. Next, we construct a deep reconstruction module based on the pyramid model, which realizes adaptive feature enhancement at different resolutions by applying the attention mechanism at different scales. We jointly optimize the sampling network and the reconstruction network, and the model obtained by training this network is able to adapt to arbitrary sampling ratios. Testing results across different datasets demonstrate that our proposed compressive sensing reconstruction network exhibits rapid operational speed while ensuring the high quality of image reconstruction.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105413"},"PeriodicalIF":2.9,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144338891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AED-YOLO11: A small object detection model based on YOLO11","authors":"Xuejian Gong ,&nbsp;Jiong Yu ,&nbsp;Huayi Zhang ,&nbsp;Xinsheng Dong","doi":"10.1016/j.dsp.2025.105411","DOIUrl":"10.1016/j.dsp.2025.105411","url":null,"abstract":"<div><div>Small object detection suffers from inherent challenges including noise susceptibility, frequent occlusions, low spatial feature saliency, and imbalanced data distribution. While You Only Look Once 11 (YOLO11) maintains real-time processing capabilities, its detection efficacy on small objects is compromised by insufficient frequency-domain analysis and redundant computational operations in shallow network layers. To overcome these challenges, this study introduces Adaptive Efficient and Dynamic-YOLO11 (AED-YOLO11), a novel detection framework built upon the YOLO11 architecture with specialized enhancements for small object recognition. Specifically, the model introduces the following innovations: First, the Adaptive Frequency Domain Aggregation (AFDA) module dynamically aggregates features using frequency-domain information and channel-wise weighting, resolving frequency inconsistencies in small object images. Second, the Efficient Attention Compression (EAC) module significantly reduces computational costs by compressing channel dimensions and fusing features, thereby improving feature extraction capabilities. Third, the Dynamic Upsampling (DySample) module enhances spatial transformation capabilities through dynamic sampling of input feature maps. Finally, the Wise-IoU(WIoU) loss function is applied to improve detection performance on low-quality samples. Additionally, the detection head structure is optimized to better suit small object detection needs. Collectively, these improvements enhance the model's accuracy and computational efficiency, demonstrating superior performance in complex scenarios. Benchmark tests on VisDrone2019 indicate AED-YOLO11 yields a 4.2% mAP enhancement over baseline approaches while surpassing existing YOLO-series models in small object recognition tasks.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105411"},"PeriodicalIF":2.9,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144329998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DT-Retinex: low-light enhancement network based on diffuse denoising and light enhancement","authors":"Wenjuan Gu ,&nbsp;Xin Li ,&nbsp;Yuhanke Hu ,&nbsp;Junxiang Peng ,&nbsp;Xiaobao Liu","doi":"10.1016/j.dsp.2025.105416","DOIUrl":"10.1016/j.dsp.2025.105416","url":null,"abstract":"<div><div>Low-light images often suffer from insufficient brightness, blurred details, and noise interference, which degrade visual quality and reduce the accuracy of computer vision tasks. To address these challenges, this paper proposes a low-light image enhancement model named DT-Retinex. The method improves image quality through three stages: image decomposition, reflectance denoising, and illumination enhancement. First, the decomposition network decouples the input image into reflectance and illumination components while preserving structural features. Then, a diffusion model is introduced to progressively denoise the reflectance component, with a customized denoising loss designed to enhance detail restoration. Finally, DT-Retinex adopts an encoder-decoder architecture for illumination enhancement: the encoder extracts multi-level features and leverages the LIT module to model global illumination, while the decoder incorporates CBAM attention to emphasize key regions and adaptively adjust lighting information during spatial reconstruction. Experimental results show that DT-Retinex outperforms existing methods on several benchmark datasets, achieving excellent performance on PSNR, SSIM, and LPIPS, as well as better perceptual naturalness and consistency under no-reference metrics such as NIQE and BRISQUE. Overall, DT-Retinex provides a robust and high-quality solution for low-light image enhancement tasks.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105416"},"PeriodicalIF":2.9,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144330578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-time instance segmentation algorithm based on mask activation and feature enhancement","authors":"Linlin Ma ,&nbsp;Jianwei Zhang ,&nbsp;Zengyu Cai ,&nbsp;Jianxin Ma","doi":"10.1016/j.dsp.2025.105402","DOIUrl":"10.1016/j.dsp.2025.105402","url":null,"abstract":"<div><div>With the widespread deployment of the Internet of Things, the demand for real-time environmental perception has become increasingly urgent. In this context, instance segmentation technology has emerged as a pixel-level scene perception method, garnering significant attention. This paper proposes a novel and efficient instance segmentation network designed for precise scene perception. In the decoding stage, we design a mask activation module to construct multi-layer weight matrices, with each layer directly activating a mask region of an instance, thereby achieving simplicity and efficiency. During the feature enhancement stage, we introduce two crucial modules to improve performance. Firstly, the global feature perception module models global dependencies through the self-attention mechanism, extending the network's receptive field. Secondly, the foreground feature capture module employs parallel convolutional kernels of various shapes and sizes to comprehensively explore foreground instance information. Experimental verification on the MS-COCO dataset demonstrates that our method achieves a better balance between accuracy and speed, and has potential in practical applications.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105402"},"PeriodicalIF":2.9,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144291281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual dynamic kernel filtering: Accurate time-frequency representation, reconstruction, and denoising","authors":"Skander Bensegueni ,&nbsp;Samir Brahim Belhaouari ,&nbsp;Yunis Carreon Kahalan","doi":"10.1016/j.dsp.2025.105407","DOIUrl":"10.1016/j.dsp.2025.105407","url":null,"abstract":"<div><div>Time-frequency analysis plays a critical role in characterizing non-stationary signals such as electrocardiograms (ECG), where both spectral and temporal details are paramount. In this study, we introduce Dual Dynamic Kernel Filtering (2DKF) for time-frequency decomposition, emphasizing how kernel selection influences signal representation, reconstruction accuracy, and overall filtering performance. To overcome the limitations associated with signal-dependent single-kernel methods, we propose an innovative Dual hybrid kernel strategy that adaptively integrates multiple kernel functions to capture a wide array of signal characteristics. This approach significantly improves temporal alignment via Dynamic Time Warping (DTW), robustly preserves signal distributions as evidenced by quantile-quantile (QQ) plot analyses, and maintains high frequency fidelity during the filtering process. Extensive experimental comparisons against traditional discrete wavelet transform (DWT) and S-transform filtering, conducted under varying noise conditions, including synthetic noisy ECG with white noise, colored noise (brown and pink), and naturally noisy ECG, demonstrate that our dual hybrid kernel method substantially enhances robustness and consistency in signal reconstruction. Furthermore, we compare our approach with Recursive Multikernel Filtering (RMKF) technique for a benchmark nonlinear signal corrupted by structured noise, alongside wavelet and S-transform techniques. Evaluation metrics, including normalized mean square error (nMSE), root mean square error (RMSE) and correlation coefficients, confirm the superior performance of the proposed approach. These promising results underscore the potential of our method as a powerful tool for the time-frequency analysis of non-stationary signals, with significant implications for advanced ECG signal processing and other biomedical applications.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105407"},"PeriodicalIF":2.9,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144321447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Event-triggered state estimation for networked control systems with silent packet loss and coloured noise","authors":"Chang Zhao ,&nbsp;Yamin Wang ,&nbsp;Ka-Wai Kwok ,&nbsp;Xiaoling Liu","doi":"10.1016/j.dsp.2025.105395","DOIUrl":"10.1016/j.dsp.2025.105395","url":null,"abstract":"<div><div>This paper focuses on jointly designing a scheduler, detector, and estimator for networked control systems with silent packet loss (SPL) and coloured noise. A truncated Gaussian distribution emerges in the state estimator calculating process due to the event-triggered scheduling mechanism. Unfortunately, this distribution leads to the absence of an analytical expression in the derivation process, necessitating approximating the truncated Gaussian distribution as a Gaussian distribution within the design of the optimal estimator (OE). To overcome this issue, this paper implements a stochastic event-triggered scheduling mechanism. Moreover, a detector is devised to identify packet loss occurrences, thereby improving the estimation performance. Built upon the framework, an OE estimator is formulated. Then, a lower bound is established for the communication rate, and a necessary condition is obtained for the stability of the OE estimator in stable and unstable systems. In the end, numerical examples are provided to verify the effectiveness of theoretical results.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105395"},"PeriodicalIF":2.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144312477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automatic modulation recognition based on sample-transferable and branch-scalable method for signals in complex multipath channel","authors":"Yitong Lu,&nbsp;Shujuan Hou,&nbsp;Shiyi Yuan,&nbsp;Qin Zhang,&nbsp;Yazhe He,&nbsp;Shouzhi Wang","doi":"10.1016/j.dsp.2025.105406","DOIUrl":"10.1016/j.dsp.2025.105406","url":null,"abstract":"<div><div>At present, there are a large number of mature deep learning related studies on automatic modulation recognition (AMR) for signals in the additive white Gaussian noise (AWGN) or fixed multipath channel. However, in actual communication environments, the AMR method is required to have strong generalization ability due to the complexity and variability of multipath channels. Thus, we propose a sample-transferable and branch-scalable method suitable for signals in different multipath channels. According to the generation principle of multipath signals, we first estimate the multipath signals based on the direction of arrival (DOA) estimation algorithm to obtain characteristic parameters such as the number of paths and the direction of arrival. Then we decompose the multipath signals into multi-branch single-path signals using the estimation results. On this basis, we propose a multi-branch neural network trained with signals in the AWGN channel, with the decomposed multi-branch single-path signals serving as inputs. Hence, sample transfer from the training signals in the AWGN channel to the test signals in the multipath channel can be realized, significantly improving the generalization ability of the network. Moreover, we introduce the attention mechanism module to perform feature-level fusion on multi-branch signals, and use multipath signals to obtain additional recognition gain compared to single-path signals. In response to the uncertainty of multipath number in complex multipath channel environments, we propose a branch-scalable dynamic neural network (BSDNN) with novel “dual-branch training, multi-branch recognition”, and realize the recognition of multipath signals with arbitrary path number using the network structure trained with dual-branch signals. The experimental results show that our proposed BSDNN trained with the dual-branch signals in the AWGN channel can successfully transfer to modulation recognition of multipath signals with any number of paths. Furthermore, the method exhibits advantages in terms of lightweight design, with fewer network parameters and training time.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105406"},"PeriodicalIF":2.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144312884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design of linear-phase IIR integrators with maximally-flat and Chebyshev magnitude responses","authors":"Ivan Krstić ,&nbsp;Goran Stančić ,&nbsp;Jasna Radulović","doi":"10.1016/j.dsp.2025.105400","DOIUrl":"10.1016/j.dsp.2025.105400","url":null,"abstract":"<div><div>This paper proposes two methods for designing linear-phase infinite impulse response integrators. The first method, referred to as the maximally-flat one, imposes flatness conditions on the frequency response error function, leading to a system of linear equations that have to be solved to determine unknown coefficients. Furthermore, a relation is established between the proposed maximally-flat integrators and existing integer-order linear-phase integrators derived using the algebraic polynomial-based quadrature rules, demonstrating that the latter represent special cases of the proposed integrators. The second method, referred to as the optimal one, minimizes the complex frequency response error function in the weighted Chebyshev sense, which is achieved by an efficient exchange algorithm that exhibits rapid convergence. The proposed linear-phase integrators are also compared with several existing linear- and nearly linear-phase integrators.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105400"},"PeriodicalIF":2.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144280801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Minimization of pseudo fountain penalty for sparse signal recovery","authors":"Zhihua Li ,&nbsp;Feixiang Zhang ,&nbsp;Ning Yu","doi":"10.1016/j.dsp.2025.105404","DOIUrl":"10.1016/j.dsp.2025.105404","url":null,"abstract":"<div><div>In this paper, we propose a novel Pseudo-fountain (PF) penalty that builds upon and extends compressed sensing (CS) theory. The PF penalty optimizes dual parameters in coordination, enhancing its adaptability to the sparsity of signals. Meanwhile, leveraging the renowned RIP theory, we establish explicit conditions for the exact and robust recovery of signals. Additionally, we develop a Difference of Convex Algorithm-PF (DCA-PF) tailored for the constrained sparse signal recovery model formulated in this work. The experimental results demonstrate that the PF penalty outperforms its counterparts in terms of robustness, stability, and sparsity for sparse signal recovery.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105404"},"PeriodicalIF":2.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144297769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 novel stochastic model for the steady-state performance of norm-penalized adaptive algorithms","authors":"Lucas P.R. da Silva ,&nbsp;Fabio A.A. Andrade ,&nbsp;Milena F. Pinto ,&nbsp;Gilson A. Giraldi ,&nbsp;Diego Barreto Haddad","doi":"10.1016/j.dsp.2025.105403","DOIUrl":"10.1016/j.dsp.2025.105403","url":null,"abstract":"<div><div>This paper proposes a new model to estimate the asymptotic performance of adaptive algorithms with norm penalization of the adaptive coefficient vector. The attraction-to-zero term is modeled as a piecewise linear function, allowing the proposed approach to approximate, with arbitrary precision, the behavior of multiple algorithms from the literature. Assuming a white input signal, it is possible to derive a general model capable of predicting the algorithm's asymptotic performance in terms of mean square deviation. The closed-form expression obtained for the mean square deviation is then approximated using heuristics, allowing the optimal value of the parameter regulating the norm penalization to also be determined through a closed-form formula. The derived formulas were extensively tested and validated through simulations, demonstrating good accuracy, with a maximum error of 0.17 dB between the theoretical and simulated values.</div></div>","PeriodicalId":51011,"journal":{"name":"Digital Signal Processing","volume":"166 ","pages":"Article 105403"},"PeriodicalIF":2.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144297768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}