Deep Unrolling Networks (DUNs) integrate classical optimization recovery problems in Compressed Sensing (CS) with sophisticated deep learning network architectures, leading to substantial breakthroughs. However, prevailing DUNs generally face challenges concerning solidified gradient descent step size strategies, inadequate feature extraction within the iterative stage and limited information interaction between iterative stages. To overcome these obstacles, we propose SCU-Net, a channel-focused unrolling network inspired by the renowned spectral projected gradient optimization algorithm. In particular, we tailore two pivotal components, Barzilai-Borwein-gradient Descent Optimizer (BBDO) and Channel-guided Cross-attention Reconstruction Module (CCRM), to collaboratively undertake the reconstruction task. BBDO leverages a gradient calculation strategy based on BB step size to enhance data fidelity optimization, while CCRM addresses the intricate mapping issue associated with sparse induction, encompassing customized functionalities from Adaptive Channel Interaction Layer (ACIL) and Spatially Augmented Channel-aware Unit (SACU). Among them, ACIL amalgamates convolution operations and channel attention mechanisms to achieve meticulous information screening alongside efficient feature enhancement. SACU introduces dual reinforcement variables to bolster information exchange across different iterative stages, coupled with the optimization of cross-attention to facilitate the modeling of long-distance dependencies. Extensive experiments in both image CS and magnetic resonance imaging exhibit that our SCU-Net manifests superior performance, surpassing state-of-the-art methods.

Weakly supervised anomaly detection methods (WADMs) can effectively utilize incomplete label data to address the issue of imbalanced samples, thereby reducing the reliance on the quantity of labeled data and demonstrating superior anomaly detection capabilities in network and service management. However, when confronted with significant label noise and missing labels, existing WADMs still struggle to adequately extract the deep feature information of samples, leading to a decline in model performance. To tackle this challenge, we propose a self-contrastive enhanced weakly supervised anomaly detection framework, called SEAD-Net, which enhances the feature representation capability of data in the model’s feature space, thereby improving the accuracy and robustness of anomaly detection. Specifically, we first design a personalized data enhancement module that augments data representation by applying various transformations to the raw data. Subsequently, a self-contrastive enhanced learning module is introduced to impose hybrid constraints on the augmented samples, constructing the overall distribution structure while learning deep sample feature spaces under complex scenario disturbances. Finally, we extract contrastive enhancement features within the deep sample feature space and perform probabilistic generation to enable effective decision-making via an anomaly probability generation module. Experimental results on a series of public benchmark datasets demonstrate that our SEAD-Net outperforms the second-best WADM by 5.95% in average AUC-ROC and 16.28% in average AUC-PR.

Deep Unfolding Networks (DUNs), with their outstanding performance and partial interpretability, have revitalized the field of pan-sharpening. However, the current DUNs for pan-sharpening rely entirely on implicit deep priors, ignoring the intrinsic physical prior knowledge of multispectral image (MS) and panchromatic image (PAN) to guide the reconstruction process. Moreover, these methods often depend on single-scale prior features, failing to adequately capture multiscale information, resulting in spatial and spectral distortions in detail. In this paper, we introduce a spatial-spectral prior-aware framework for pan-sharpening, called SSPF, which formulates a constrained minimization problem integrating MS and PAN prior knowledge based on spatial and spectral domains. We further develop SSPF into a lightweight deep unfolding network, called SSUN-Net, which provides more efficient prior feature extraction and requires less computational cost. Additionally, we augment SSUN-Net's capabilities by integrating a customized multi-scale prior structure (MPS). MPS imposes constraints on the solution space at various scales through regularization, which markedly enhances the reconstruction of intricate details. Extensive experiments demonstrate the significant advantages of our proposed SSUN-Net over the current SOTA methods.

Despite advancements in Computer-Aided-Design (CAD) generation, direct generation of complex Boundary Representation (B-rep) CAD models remains challenging. The difficulty arises from the parametric nature of B-rep data, complicating the encoding and generation of its geometric and topological information. In this paper, we introduce BrepGiff, a lightweight generation approach for high-quality and complex B-rep based on 3D Graph Diffusion. First, we transfer B-rep models into 3D graphs representation. Specifically, BrepGiff extracts and integrates topological and geometric features to construct a 3D graph where nodes correspond to face centroids in 3D space, preserving adjacency and degree information. Geometric features are derived by sampling points in the UV domain and extracting face and edge features. BrepGiff then applies Graph Attention Network (GAT) to enforce topological constraints from local to global during the degree-guided diffusion process. With the 3D graph representation and diffusion process, BrepGiff significantly reduces the computational cost and improves the quality, thus achieving lightweight generation of complex models. Experiments show that BrepGiff can generate complex B-rep models (>100 faces) using only 2 RTX4090 GPUs, achieving state-of-the-art performance in B-rep generation.

Deep Unfolding Networks (DUNs) have risen to prominence due to their interpretability and superior performance for image Compressive Sensing (CS). However, existing DUNs still face significant issues, such as the insufficient representation capability of single-scale image information during the iterative reconstruction phase and loss of feature information, which fundamentally limit the further enhancement of image CS performance. In this paper, we propose Homotopy Unfolding Network (HUNet) for image CS, which enables phase-by-phase reconstruction of images along homotopy path. Specifically, each iteration step of the traditional homotopy algorithm is mapped to a Multi-scale Homotopy Iterative Module (MHIM), which includes U-shaped stacked window-based Transformer blocks capable of efficient feature extraction. Within the MHIM, we design the deep homotopy continuation strategy to ensure the interpretability of the homotopy algorithm and facilitate feature learning. Additionally, we introduce a dual-path feature fusion module to mitigate the loss of high-dimensional feature information during the transmission between iterative phases, thereby maximizing the preservation of details in the reconstructed image. Extensive experiments indicate that HUNet achieves superior image reconstruction results compared to existing state-of-the-art methods. The source code is available at https://github.com/ICSResearch/HUNet.

Cyber-Physical Systems (CPS) consist of various physical devices and have a significant impact on both industry and the economy. In CPS, controllers or control centers typically send continuous commands to actuators. Due to the limited computational and storage resources of CPS devices, existing Message Authentication Code and digital signature algorithms face challenges in terms of being non-interactive, lightweight, and continuous on the actuator side. To address this issue, we propose a lightweight continuous message authentication (LCMA) scheme that ensures the continuity of signatures through a mechanism involving random numbers and incrementing sub-private keys. The multidimensional Bloom filters is designed and used togather with precomputed verification points generated by Key Generation Center to enable non-interactive verification. Additionally, we design a redundancy-free hash tree for CPS to reduce information redundancy. The proposed solution requires only hash operations and Bloom filter queries for verification, achieving sufficient lightweight performance. Finally, we provide a formal security proof for the proposed scheme and simulate it on the ESP32 platform. The results show that the implementation times for signature generation and verification are 0.779 milliseconds and 0.792 milliseconds, respectively. Compared to other non-interactive lightweight message authentication schemes, the proposed LCMA is more suitable for CPS devices.

Industrial control systems (ICSs) are becoming increasingly interconnected, rendering them susceptible to cyber attacks. Timely detection of anomalies in encrypted data flows is crucial for ensuring the reliability and security of ICS. Although deep learning-based anomaly detection methods have made significant strides, their implementation in point-by-point mapping paradigms often necessitates a tradeoff between feature representation and computational efficiency, especially in resource-constrained environments. In addition, these methods face challenges with class imbalance and limited labeled anomaly data. To address these challenges, we propose a frequency-domain anomaly detection (FreAD) framework specifically tailored for encrypted data flows in ICS. FreAD utilizes a frequency-domain feature fusion encoding module to capture global temporal dependencies. An anomaly scoring network integrates a small amount of labeled data along with pseudolabeled data generated from the modified Z-score algorithm, effectively addressing class imbalance. Furthermore, the frequency-domain deviation prior module can alleviate the contraction of interclass distances during unsupervised training. Extensive experiments demonstrate that FreAD significantly surpasses other state-of-the-art anomaly detection algorithms.

Existing unfolding-based compressive imaging approaches always suffer from certain issues, including inefficient feature extraction and information loss during iterative reconstruction phases, which become particularly evident at low sampling ratios, i.e., significant detail degradation and distortion in reconstructed images. To mitigate these challenges, we propose USB-Net, a deep unfolding method inspired by the renowned Split Bregman algorithm and multi-phase feature integration strategy, for compressive imaging reconstruction. Specifically, we use a customized Depthwise Attention Block as a fundamental block for feature extraction, but also to address the sparse induction-related splitting operator within Split Bregman method. Based on this, we introduce three Auxiliary Iteration Modules:${\mathrm {X}}^{(k)}$,${\mathrm {D}}^{(k)}$, and${\mathrm {B}}^{(k)}$to reinforce the effectiveness of Split Bregman’s decomposition strategy for problem breakdown and Bregman iterations. Moreover, we introduce two categories of Iterative Fusion Modules to seamlessly harmonize and integrate insights across iterative reconstruction phases, enhancing the utilization of crucial features, such as edge information and textures. In general, USB-Net can fully harness the advantages of traditional Split Bregman approach, manipulating multi-phase iterative insights to enhance feature extraction, optimize data fidelity, and achieve high-quality image reconstruction. Extensive experiments show that USB-Net significantly outperforms current state-of-the-art methods on image compressive sensing, CS-magnetic resonance imaging, and snapshot compressive imaging tasks, demonstrating superior generalizability. Our code is available at USB-Net.

Deep Learning (DL)-based weakly supervised anomaly detection methods enhance the security and performance of communication and networks by promptly identifying and addressing anomalies within imbalanced samples, thus ensuring reliable communication and smooth network operations. However, existing DL-based methods often overly emphasize the local feature representations of samples, thereby neglecting the long-range dependencies and the prior knowledge of the samples, which imposes potential limitations on anomaly detection with a limited number of abnormal samples. To mitigate these challenges, we propose a Transformer deviation network for weakly supervised anomaly detection, called TFD-Net, which can effectively leverage the interdependencies and data priors of samples, yielding enhanced anomaly detection performance. Specifically, we first use a Transformer-based feature extraction module that proficiently captures the dependencies of global features in the samples. Subsequently, TFD-Net employs an anomaly score generation module to obtain corresponding anomaly scores. Finally, we introduce an innovative loss function for TFD-Net, named Transformer Deviation Loss Function (TFD-Loss), which can adequately incorporate prior knowledge of samples into the network training process, addressing the issue of imbalanced samples, and thereby enhancing the detection efficiency. Experimental results on public benchmark datasets demonstrate that TFD-Net substantially outperforms other DL-based methods in weakly supervised anomaly detection task.

Efficient real-time communication in Time-Sensitive Networking (TSN) relies on precise flow scheduling to meet stringent latency and reliability requirements. However, under the Cyclic Queuing and Forwarding (CQF) model, existing scheduling algorithms face challenges in resource allocation efficiency and the scheduling of unstable flows, leading to inconsistent performance across complex network environments. To address these challenges, firstly, this paper proposes a Formal Scheduling Architecture for CQF (CQF-FSA), which rigorously defines key scheduling elements and constraints, providing a basic, consistent, and reusable architecture for scheduling algorithms across diverse network environments; Secondly, based on CQF-FSA, we propose an optimized scheduling algorithm, NTOS (Neuro-Tabu Optimized Scheduler), which combines the global exploration capabilities of NEAT (NeuroEvolution of Augmenting Topologies) with the local optimization efficiency of Tabu search. NTOS effectively overcomes the limitations of existing methods by optimizing resource utilization and reducing scheduling conflicts; Finally experimental results demonstrate that NTOS improves the scheduling success rate by an average of 34.5% over the NV algorithm and 3.23% over the state-of-the-art MSS algorithm across various network topologies. This paper provides a highly optimized solution for CQF scheduling in TSN, significantly enhancing scheduling efficiency and scalability.

In the domain of compressive sensing (CS), deep unfolding networks (DUNs) have garnered attention for their good performance and certain degree of interpretability rooted in CS domain, achieved by marrying traditional optimization solvers with deep networks. However, current DUNs are ill-suited for the intricate task of capturing fine-grained image details, leading to perceptible distortions and blurriness in reconstructed images, particularly at low CS ratios, e.g., 0.10 and below. In this paper, we propose CPP-Net, a novel deep unfolding CS framework, inspired by the primal-dual hybrid strategy of the Chambolle and Pock Proximal Point Algorithm (CP-PPA). First, we derive three iteration submodules, X(k), V(k) and y(k), by incorporating customized deep learning modules to solve the sparse basis related proximal operator within CP-PPA. Second, we de-sign the Dual Path Fusion Block (DPFB) to adeptly extract and fuse multi-scale feature information, enhancing sensi-tivity to feature information at different scales and improving detail reconstruction. Third, we introduce the Iteration Fusion Strategy (IFS) to effectively weight the fusion of outputs from diverse reconstruction stages, maximizing the utilization of feature information and mitigating the information loss during reconstruction stages. Extensive experiments demonstrate that CPP-Net effectively reduces distortion and blurriness while preserving richer image details, outperforming current state-of-the-art methods. Codes are available at https://github.com/ICSResearch/CPP-Net.

Deep unfolding networks (DUNs), renowned for their in-terpretability and superior performance, have invigorated the realm of compressive sensing (CS). Nonetheless, existing DUNs frequently suffer from issues related to insufficient feature extraction and feature attrition during the it-erative steps. In this paper, we propose Unrolling Fixed-point Continuous Network (UFC-Net), a novel deep CS framework motivated by the traditional fixed-point contin-uous optimization algorithm. Specifically, we introduce Convolution-guided Attention Module (CAM) to serve as a critical constituent within the reconstruction phase, encompassing tailored components such as Multi-head Attention Residual Block (MARB), Auxiliary Iterative Reconstruction Block (AIRB), etc. MARB effectively integrates multi-head attention mechanisms with convolution to reinforce feature extraction, transcending the confinement of localized attributes and facilitating the apprehension of long-range correlations. Meanwhile, AIRB introduces auxiliary vari-ables, significantly bolstering the preservation of features within each iterative stage. Extensive experiments demon-strate that our proposed UFC-Net achieves remarkable per-formance both on image CS and CS-magnetic resonance imaging (CS-MRI) in contrast to state-of-the-art methods.

Deep unrolling architectures have revitalized compressive sensing (CS) by seamlessly blending deep neural networks with traditional optimization-based reconstruction algorithms. In pursuit of an efficient and deep interpretable approach, we propose LTwIST for CS problem, a novel deep unrolling framework that draws inspiration from the well-known two-step iterative shrinkage thresholding (TwIST) algorithm. LTwIST uses a trainable sensing matrix to adaptively learn structural information in images, and introduces a customized U-block architecture to solve the proximal mapping of nonlinear transformations connected with the sparsity-inducing regularizer. Specifically, each iteration recovery step of LTwIST corresponds to an iterative update step of the traditional TwIST algorithm. Moreover, the proposed method is designed to learn all the parameters end-to-end without manual tuning such as shrinkable thresholds, step sizes, etc. As a result, LTwIST obviates the need for manual parameter optimization, allows for high-quality image recovery and provides unambiguous interpretability. Moreover, our proposed LTwIST is also applicable to CS-based magnetic resonance imaging and exhibits a strong reconstruction performance. Extensive experiments on several public benchmark datasets demonstrate that the proposed LTwIST outperforms existing state-of-the-art deep CS methods by considerable margins in terms of quality evaluation metrics and visual performance. Our code is available on LTwIST.

Image compressed sensing (ICS) has been extensively applied in various imaging domains due to its capability to sample and reconstruct images at subNyquist sampling rates. The current predominant approaches in ICS, specifically pure convolutional networks (ConvNets)-based ICS methods, have demonstrated their effectiveness in capturing local features for image recovery. Simultaneously, the Transformer architecture has gained significant attention due to its capability to model global correlations among image features. Motivated by these insights, we propose a novel hybrid network for ICS, named MTC-CSNet, which effectively combines the strengths of both ConvNets and Transformer architectures in capturing local and global image features to achieve high-quality image recovery. Particularly, MTC-CSNet is a dual-path framework that consists of a ConvNets-based recovery branch and a Transformer-based recovery branch. Along the ConvNets-based recovery branch, we design a lightweight scheme to capture the local features in natural images. Meanwhile, we implement a Transformer-based recovery branch to iteratively model the global dependencies among image patches. Ultimately, the ConvNets-based and Transformer-based recovery branches collaborate through a bridging unit, facilitating the adaptive transmission and fusion of informative features for image reconstruction. Extensive experimental results demonstrate that our proposed MTC-CSNet surpasses the state-of-the-art methods on various public datasets. The code and models are publicly available at MTC-CSNet.

Deep compressive sensing (CS) has become a prevalent technique for image acquisition and reconstruction. However, existing deep learning (DL)-based CS methods often encounter challenges such as block artifacts and information loss during iterative reconstruction, particularly at low sampling rates, resulting in a reduction of reconstructed details. To address these issues, we propose NesTD-Net, an unfolding-based architecture inspired by the NESTA algorithm, designed for image CS. NesTD-Net integrates DL modules into NESTA iterations, forming a deep network that continuously iterates to minimize the ℓ1-norm CS problem, ensuring high-quality image CS. Utilizing a learned sampling matrix for measurements and an initialization module for initial estimate, NesTD-Net then introduces Iteration Sub-Modules derived from the NESTA algorithm (i.e., Yk, Zk, and Xk) during reconstruction stages to iteratively solve the ℓ1-norm CS reconstruction. Additionally, NesTD-Net incorporates a Dual-Path Deblocking Structure (DPDS) to facilitate feature information flow and mitigate block artifacts, enhancing image detail reconstruction. Furthermore, DPDS exhibits remarkable versatility and demonstrates seamless integration with other unfolding-based methods, offering the potential to enhance their performance in image reconstruction. Experimental results demonstrate that our proposed NesTD-Net achieves better performance compared to other state-of-the-art methods in terms of image quality metrics such as SSIM and PSNR, as well as visual perception on several public benchmark datasets. Our code is available at NesTD-Net.

Brain-teleoperation robot control ensures that human beings interact with telepresence mobile systems through the brain neural signals. In this study, a hierarchical robust predictive control framework consisting of a two-loop control scheme is developed to simultaneously enhance the safety, navigation, and robustness performance of electroencephalography (EEG)-based robotic systems and minimize the loss of control by the end-user. The outer loop is a model-based predictive controller to guarantee the optimal velocity evolution under various constraints. The inner loop is the integral sliding mode controller constructed by a novel integral sliding manifold and enables the velocity tracking properties under uncertainty compensation. Human-in-the-loop driving experiments are performed under different disturbances, and the results show that the proposed system offers advantages of safety, enhanced navigation performance, and stronger robustness over those conventional direct control of EEG-based robots. Therefore, brain-robot teleoperation is improved in terms of robust motion control and velocity modulation, providing insights into similar brain-controlled dynamic systems.

Driven by the widespread applications of Space-Air-Ground Integrated Networks (SAGINs) in a number of practical fields, the volume of space data grows rapidly. However, the large volume of space data in SAGINs is typically intractable to be offloaded from space to the ground under the high dynamic network topology and the stochastic data arrivals. Furthermore, most nodes in SAGINs are battery-powered and energy-constrained, thereby implying that energy consumption becomes one major bottleneck for data offloading. Towards this end, this paper studies online joint data offloading and power control in SAGINs to maximize long-term time-averaged data offloaded amount under the constraints of average energy consumption. First, we propose a novelty Two-timescale Time-Expanded Graph (TTEG) to characterize the rapid change of the network topology in large-timescale slots and capture the stochastic data arrivals in small-timescale slots. Based the TTEG model, we formulate a stochastic optimization problem and transform it into a series of per-time-slot subproblems to obtain an efficient online solution. Through theoretical analyses, we show that the performance gap with optimal solution is bounded. Finally, extensive simulations demonstrate that the maximum performance gap of our proposed online solution to the optimal solution is less than 2% in a low computation cost.

Predicting a high quality depth map from a single image is a challenging task, because it exists infinite possibility to project a 2D scene to the corresponding 3D scene. Recently, some studies introduced multi-head attention (MHA) modules to perform long-range interaction, which have shown significant progress in regressing the depth maps. The main functions of MHA can be loosely summarized to capture long-distance information and report the attention map by the relationship between pixels. However, due to the quadratic complexity of MHA, these methods can not leverage MHA to compute depth features in high resolution with an appropriate computational complexity. In this paper, we exploit a depth-wise convolution to obtain long-range information, and propose a novel trap attention, which sets some traps on the extended space for each pixel, and forms the attention mechanism by the feature retention ratio of convolution window, resulting in that the quadratic computational complexity can be converted to linear form. Then we build an encoder-decoder trap depth estimation network, which introduces a vision transformer as the encoder, and uses the trap attention to estimate the depth from single image in the decoder. Extensive experimental results demonstrate that our proposed network can outperform the state-of-the-art methods in monocular depth estimation on datasets NYU Depth-v2 and KITTI, with significantly reduced number of parameters. Code is available at: https://github.com/ICSResearch/TrapAttention.

Block compressive sensing (CS) is a well-known signal acquisition and reconstruction paradigm with widespread application prospects in science, engineering, and cybernetic systems. However, state-of-the-art block-based image CS (BCS) methods generally suffer from two issues. The sparsifying domain and the sensing matrices widely used for image acquisition are not data driven and, thus, both the features of the image and the relationships among subblock images are ignored. Moreover, it requires to address a high-dimensional optimization problem with extensive computational complexity for image reconstruction. In this article, we provide a deep learning (DL) strategy for BCS, called AutoBCS, which automatically takes the prior knowledge of images into account in the acquisition step and establishes a reconstruction model for performing fast image reconstruction. More precisely, we present a learning-based sensing matrix to accomplish image acquisition, thereby capturing and preserving more image characteristics than those captured by the existing methods. In addition, we build a noniterative reconstruction network, which provides an end-to-end BCS reconstruction framework to maximize image reconstruction efficiency. Furthermore, we investigate comprehensive comparison studies with both traditional BCS approaches and newly developed DL methods. Compared with these approaches, our proposed AutoBCS can not only provide superior performance in terms of image quality metrics (SSIM and PSNR) and visual perception but also automatically benefit reconstruction speed.

Under false data-injection (FDI) attacks, the data of some agents are tampered with by the FDI attackers, which causes that the distributed algorithm cannot estimate the ideal unknown parameter. Due to the concealment of the malicious data tampered with by the FDI attacks, many detection algorithms against FDI attacks often have poor detection results or low detection efficiencies. To solve these problems, a conveniently distributed diffusion least-mean-square (DLMS) algorithm with cross-verification (CV) is proposed against FDI attacks. The proposed DLMS with CV (DLMS-CV) algorithm is comprised of two subsystems: one subsystem provides a detection test of agents based on the CV mechanism, while the other provides a secure distribution estimation. In the CV mechanism, a smoothness strategy is introduced, which can improve the detection performance. The convergence performance of the proposed algorithm is analyzed, and then the design method of the adaptive threshold is also formulated. In particular, the probabilities of missing alarm and false alarm are examined, and they decay exponentially to zero under sufficiently small step size. Finally, simulation experiments are provided to illustrate the effectiveness and simplicity of the proposed DLMS-CV algorithm in comparison to other algorithms against FDI attacks.

Ship detection using polarimetric SAR (PolSAR) images has recently been an active topic in the Earth observation field. There, how to detect small ships is an open and challenging issue. Within this context, we put forward a two-stage ship detection model, by which a novel ship detection method is proposed as well. Briefly, in the first stage, a suppression manipulation is adopted to suppress sea clutter, where the feature SVVSOis built on the intensity information with the orientation angle compensation (OAC). In the second stage, an enhancement manipulation is further executed to highlight ships from the suppressed sea clutter, where the features PID (polarimetric intensity difference) and NsD (nonsurface degree) are first constructed with SVVSOand a series of theoretical derivations. Then, via fusing PID and NsD together, the two-stage-based method FPAN is proposed to detect ships. To demonstrate its performance, we apply FPAN to four different L-Band PolSAR datasets. Experimental results reveal that, compared to other state-of-the-art methods, especially the DBSPCPmethod, FPAN is more effective in detecting small ships. On average, its figure-of-merit (FoM) and target-to-clutter ratio (TCR) values are, respectively 9.40% and 25.18% greater than those of DBSPCP, while the time consumption is just 58.67% of the latter.

Well-known compressed sensing (CS) is widely used in image acquisition and reconstruction. However, accurately reconstructing images from measurements at low sampling rates remains a considerable challenge. In this paper, we propose a novel Transformer-based hybrid architecture (dubbed TransCS) to achieve high-quality image CS. In the sampling module, TransCS adopts a trainable sensing matrix strategy that gains better image reconstruction by learning the structural information from the training images. In the reconstruction module, inspired by the powerful long-distance dependence modelling capacity of the Transformer, a customized iterative shrinkage-thresholding algorithm (ISTA)-based Transformer backbone that iteratively works with gradient descent and soft threshold operation is designed to model the global dependency among image subblocks. Moreover, the auxiliary convolutional neural network (CNN) is introduced to capture the local features of images. Therefore, the proposed hybrid architecture that integrates the customized ISTA-based Transformer backbone with CNN can gain high-performance reconstruction for image compressed sensing. The experimental results demonstrate that our proposed TransCS obtains superior reconstruction quality and noise robustness on several public benchmark datasets compared with other state-of-the-art methods. Our code is available on TransCS.

Ship detection of polarimetric SAR (PolSAR) imagery has attracted lots of attentions in recent years. Also, it is known that among the polarimetric channels$HH, HV$, and$VV, VV$is the most sensitive to sea clutter. Following this guidance, in this paper, a novel ship detector SVVS is first proposed via subtracting the term$C_{33}$from the total power detector SPAN. And then, the complect polarimetric covariance difference matrix [$CP$] is utilized to calculate SVVS, leading to the construction of another novel ship detector$\text{SVVS}_{CP}$. Finally, we investigate the statistical distribution of sea clutter with$\text{SVVS}_{CP}$and further develop an adaptive$\text{SVVS}_{CP}$-based C-FAR detector for ship detection. The experiment carried out on one real PolSAR imagery shows that, compared to SPAN, both SVVS and$\text{SVVS}_{CP}$hold better ship detection performances.

In this article, we investigate the scattering components of ships and find that the surface scattering may be the primary scattering for some ships, especially small ships. Meanwhile, the drawbacks of the complete polarimetric covariance difference matrix [CP] are also pointed out in theory. Based on these analyses, two new methods are then constructed to detect the ships. More specifically, the first one RsP is constructed by directly combining the similarity parameter of surface scattering Rs and the power-maximization synthesis (PMS) detector. The second one RsDVH is designed by taking advantage of four different features (i.e., Rs, double-bounce scattering, volume scattering, and helix scattering), which are all derived from the joint polarimetric information that is developed by combing the information of the polarimetric covariance matrix [C] and [CP]. Subsequently, the generalized Gamma distribution (GΓD) is found suitable for characterizing the RsDVH values of the sea clutter. At last, an adaptive constant false-alarm-rate (CFAR) detector developed from RsDVH is proposed for ship detection. To verify the effectiveness of RsP and RsDVH, four polarization synthetic aperture radar (PolSAR) imageries are tested, including one L-band UAVSAR imagery with 19 ships, two L-band AIRSAR imageries with 22 and 53 ships, respectively, and one C-band GF-3 imagery with ten ships. The experimental results show that: 1) the surface scattering is beneficial to detecting ships, especially the ships with prominent surface scatterings; 2) compared with other state-of-the-art methods, RsDVH can more effectively enhance the target-to-clutter ratio (TCR) values of small ships in the case of rough sea surface; and 3) the joint polarimetric information that is put forward and exploited for the first time in this article has a greater potential to help ship detectors improve their detection performances than the traditional polarimetric information included in [C].

In this paper, a novel hyperspectral (HS) image fusion method using matting model is presented. Matting model refers to each band of an HS image that can be decomposed into three components, i.e., alpha channel, spectral foreground, and spectral background. First, panchromatic (PAN) image is sharpened to enhance details, and the spatial information of each band of HS image is obtained by weighted least squares filtering. Different from traditional matting model based methods that PAN image is served as the alpha channel, we do the PCA transformation to PAN image and spatial information of each band to obtain the first principal component channel which is selected for the alpha channel. This processing reduces spatial distortion. Finally, HS foreground and HS background are estimated by the alpha channel, and the fused HS image is reconstructed nearly perfectly. Experiments reveal that the proposed method is superior to the state-of-the-art methods.

Consider the cluster-based wireless sensor networks (WSNs) within a set of sensor nodes which need a reliable data broadcast, most recent works ignore the broadcast cost and the cache of sensor nodes when using instantly decodable network codes (IDNC). In this paper, we firstly propose a 2C-IDNC graph by taking advantage of Cache-IDNC graph and Cost-IDNC graph. To find a maximum weight clique in 2C-IDNC graph with smaller broadcast cost, we utilize the dynamic selection to achieve a proper broadcast cost, which can ensure the decode opportunities of encoded packets in WSNs. To this end, a heuristic algorithm is proposed to reduce the complexity of computation which operates on the novel 2C-IDNC graph. The simulation results show that the broadcast cost can be indeed reduced by the 2C- IDNC scheme.