Signal Processing最新文献

Multi-scale feature fusion has been recognized as an effective strategy to boost the quality of low-light images. However, most existing methods directly extract multi-scale contextual information from severely degraded and down-sampled low-light images, resulting in a large amount of unexpected noise and degradation contaminating the learned multi-scale features. Moreover, there exist large redundant and overlapping features when directly concatenating multi-scale feature maps, which fails to consider different contributions of different scales. To conquer the above challenges, this paper presents a novel approach termed progressive Refined-Mixed Attention Network (RMANet) for low-light image enhancement. The proposed RMANet first targets a single-scale pre-enhancement and then progressively increases multi-scale spatial-channel attention fusion in a coarse-to-fine fashion. Additionally, we elaborately devise a Refined-Mixed Attention Module (RMAM) to first learn a parallel spatial-channel dominant features and then selectively integrate dominant features in the spatial and channel dimensions across multiple scales. Noticeably, our proposed RMANet is a lightweight yet flexible end-to-end framework that adapts to diverse application scenarios. Thorough experiments carried out upon three popular benchmark databases demonstrate that our approach surpasses existing methods in terms of both quantitative quality metrics and visual quality assessment. The code will be available at https://github.com/kbzhang0505/RMANet.

Recently, conformal arrays have attracted considerable interest because such arrays can provide reduced radar cross-section and increased angle coverage. In this article, we devise a robust adaptive beamforming (RAB) approach using cylindrical uniform conformal array (CUCA). Firstly, we derive the minimum variance distortionless response (MVDR) beamformer for the CUCA by utilizing the noise subspace of interference covariance matrix (ICM) and steering vector (SV) of the signal-of-interest (SOI). Subsequently, the ICM is reconstructed by estimating the noise-free covariance matrix of the CUCA outputs and the interference projection matrix. Specifically, the noise-free covariance matrix can be regarded as multiple low-rank covariance matrices, and each low-rank matrix is reconstructed by formulating a nuclear norm minimization (NNM) problem. With the reconstructed covariance matrix, the 2-D DOAs of sources are determined by employing 2-D MUSIC spectrum to form the interference projection matrix. In addition, the SOI SV is estimated by solving a quadratically constrained quadratic programming (QCQP) problem. Numerical results demonstrate that the proposed approach is obviously superior to the existing RAB techniques.

High-quality underwater images can intuitively reflect the most realistic underwater conditions, guiding for underwater environmental monitoring and resource exploration strongly. But when factors like light absorption affect underwater optical imaging, it has been found that poor visibility and blurred texture details occur in acquired images, posing challenges for the identification and detection of underwater targets. To obtain natural images, an enhancement algorithm is proposed based on multi-scale layer decomposition and fusion. The algorithm employs different strategies to recover image attenuation information from both local and global perspectives, generating two complementary preprocessed fusion inputs. For fusion input 1, operations are conducted in the RGB color space. Initially, the mean proportion of each color channel is used to identify the attenuated color channel. Then, a local compensation strategy is adaptively applied to restore the pixel intensity of the attenuated color channel. Finally, a statistical color correction method is used to eliminate color cast in the image. Fusion input 2 involves two processing stages. In the Lab color space, the algorithm uses the grayscale information to reduce the deviation in the mean values of channels a and b globally. The local mean information of the component L enhances detail textures. In the RGB color space, linear stretching is applied to correct color deviations. To fuse the structural features of two complementary preprocessed inputs and avoid interference between signals from different layers, the color channels of fused input image are first decomposed into muti-scale structural layers based on structural priors. Then, the image enhancement is achieved through layer-by-layer fusion of the corresponding color channels of the two inputs. By testing and analyzing with the, it was found that the proposed method can improve the clarity of attenuated images in various underwater scenarios of UIEBD and RUIE datasets effectively, enhancing image detail and texture richness, increases contrast, and achieving natural and comfortable visual quality. Compared with the quantitative metrics of 14 other algorithms, the proposed algorithm shows an average score improvement of 10.14, 90.48, and 2.06, respectively, in metrics AG (average gradient), EI (edge intensity), and NIQE (natural image quality evaluator). In the RUIE dataset, it shows an average score improvement of 10.21, 94.76, and 1.86, respectively.

This paper investigates the problem of distributed target detection in the presence of interference and Gaussian noise, where the target signal and interference are assumed to lie in different deterministic subspaces. Building upon this assumption, we propose several adaptive detectors resorting to the gradient criterion tailored for homogeneous environment and partially homogeneous environment. Simulation results indicate that the proposed gradient-based detectors outperform their competitors in some scenarios. Furthermore, all of these Gradient-based detectors exhibit the constant false alarm rate (CFAR) property.

Most previous works on sound event detection (SED) are based on binary hard labels of sound events, leaving other scales of information underexplored. To address this problem, we introduce multiple granularities of knowledge into the system to perform hierarchical acoustic information fusion for SED. Specifically, we present an interactive dual-conformer (IDC) module to adaptively fuse the medium-grained and fine-grained acoustic information based on the hard and soft labels of sound events. In addition, we propose a scene-dependent mask estimator (SDME) module to extract the coarse-grained information from acoustic scenes, introducing the scene-event relationships into the SED system. Experimental results show that the proposed IDC and SDME modules efficiently fuse the acoustic information at different scales and therefore further improve the SED performance. The proposed system achieved Top 1 performance in DCASE 2023 Challenge Task 4B.

Federated Learning is a distributed machine learning paradigm that enables multiple participants to collaboratively train models without compromising the privacy of any party involved. Currently, vertical federated learning based on XGBoost is widely used in the industry due to its interpretability. However, existing vertical federated XGBoost algorithms either lack sufficient security, exhibit low efficiency, or struggle to adapt to large-scale datasets. To address these issues, we propose EVFeX, an efficient vertical federated XGBoost algorithm based on optimized secure matrix multiplication, which eliminates the need for time-consuming homomorphic encryption and achieves a level of security equivalent to encryption. It greatly enhances efficiency and remains unaffected by data volume. The proposed algorithm is compared with three state-of-the-art algorithms on three datasets, demonstrating its superior efficiency and uncompromised accuracy. We also provide theoretical analyses of the algorithm’s privacy and conduct a comparative analysis of privacy, efficiency, and accuracy with related algorithms.

The performance of traditional algorithms for spherical angle-of-arrival (AOA) source localization will be significantly degraded when there are outliers in the angle measurements. By using the symmetric $\alpha$-stable ($S\alpha S$) distribution to describe the measurement noise containing outliers and constructing the cost function using the $l_p$-norm, we propose a robust algorithm for spherical AOA source localization: the spherical iteratively reweighted pseudolinear estimator (SIRPLE). The SIRPLE is similar to the iteratively reweighted least squares (IRLS), with the difference that a homogeneous least squares (HLS) problem is solved in each iteration. The SIRPLE suffers from bias problems owing to the nature of the pseudolinear estimators. To overcome this problem, the instrumental variable (IV) method is introduced and the spherical iteratively reweighted instrumental variable estimator (SIRIVE) is proposed. Theoretical analysis shows that the SIRIVE is asymptotically unbiased and it can achieve the theoretical error covariance of the constrained least $l_p$-norm estimation. Extensive simulation analyses demonstrate the better performance of the SIRIVE compared to the conventional spherical AOA source localization methods and the SIRPLE under $S\alpha S$ noise environment. The performance of the SIRIVE is similar to that of the Nelder–Mead algorithm (NM), but the SIRIVE are computationally more efficient. In addition, the SIRIVE is nearly unbiased and the root mean square error (RMSE) performance is close to the Cramér–Rao lower bound (CRLB).

This paper considers the problem of multitarget tracking (MT) under situations where sensors have limited resolution, which leads to the presence of merged measurements (MMs). In general, an algorithm for MT under MMs can be derived by extending its standard MT counterpart which assumes that each measurement can come from at most one target. However, such an extension is by no means trivial due to the fact that one must consider data association between target groups to measurements, which results in exponential computational increasing along with the number of targets. In order to address such a difficulty, this paper proposes to adopt the message passing (MP) algorithm, and a new factor graph is constructed for MT under MMs. Then the sum–product algorithm (SPA) and max-sum algorithm (MSA) is jointly exploited for belief propagation, where the SPA is adopted for calculating the messages used for prediction and update, and the MSA is employed for efficiently perform data association. The analytical Gaussian mixture (GM) implementation is also devised for the proposed algorithm. Computational burden analyses show that the computational complexity of proposed algorithm scales linearly with respect to the number of targets and measurements. The performance of proposed algorithm is demonstrated via simulations.

In compressive sensing, Orthogonal Matching Pursuit (OMP) is a greedy algorithm used for recovering sparse signals from their incomplete linear measurements. Conventionally, the OMP algorithm relies on both the measurement matrix and the measurement signal to reconstruct sparse signals. A sensing matrix can be designed to have a small mutual coherence with respect to (w.r.t.) the measurement matrix, which is used to boost the performance of the OMP algorithm in sparse signal reconstruction. Nevertheless, sensing matrices designed by current methods are vulnerable to measurement noises. In this paper, we begin by examining the underlying cause of the non-robustness to measurement noises exhibited by these sensing matrices. Subsequently, we propose a novel approach to design a robust sensing matrix capable of withstanding the influence of measurement noises. Finally, we conduct numerical simulations to demonstrate the effectiveness and robustness of the sensing matrix designed by the proposed method.

In this paper, we propose a novel dual-head ensemble Transformer (DHET) algorithm for the classification of signals with time–frequency features such as bearing vibration signals. The DHET model employs a dual-input time–frequency architecture, integrating a 1D Transformer model and a 2D Vision Transformer model to capture the spatial and time–frequency features. By utilizing data from both the time and time–frequency domains, the proposed algorithm broadens its feature extraction capabilities and enhances the model’s capacity for generalization. In our DHET structure, the original Transformer model leverages self-attention mechanisms to consider relationships among signal input segmentations, which makes it effective at capturing long-range dependencies in signal data, while the Vision Transformer model takes 2D images as input and creates the image patches for embedding and each patch is linearly embedded into a flat vector and treated as a ‘token,’ then the ‘tokens’ are processed by the Transformer layers to learn global contextual representations, enabling the model to perform signal classification task. This integration notably enhances the performance and capability of the model. Our DHET is especially effective for rolling bearing fault diagnosis. The simulation results show that the proposed DHET has higher classification accuracy for bearing fault diagnosis and outperforms CNN-based methods.