Pavement crack detection is a fundamental task for maintaining the stability and sustainability of transportation infrastructure systems. Previous convolutional neural network (CNN) and Transformer-based methods have achieved high accuracy in general crack detection. However, processing slender crack images collected by uncrewed aerial vehicles (UAVs) and inspection vehicles remains extremely challenging. To this end, this paper proposes a pavement crack evaluation framework using a hybrid CNN-Mamba network and an air-ground platform. Firstly, the framework employs an air-ground platform composed of UAVs and vehicles to collect pavement crack data and perform preprocessing. Secondly, a hybrid CNN-Mamba network with wavelet transform, named WTCMamba, is proposed to achieve crack segmentation. Specifically, the network consists of a lightweight encoder composed of multiple dual convolution (DC) modules, a Mamba decoder, and a contextual spatial feature propagation (CSFP) module. The key innovation of the network is that the Mamba decoder comprises multiple wavelet-guided Mamba (WGM) modules, which introduce wavelet transform to convert feature channels into the frequency domain and fuse features in the global space. Finally, a grid-based quantitative risk evaluation method and a correlation analysis method are employed to analyze the overall risk of crack clusters and the morphological features of individual cracks. In addition, WTCMamba is deployed and tested on edge computing devices, achieving 35.63 FPS. Extensive experiments demonstrate that WTCMamba achieves excellent crack detection performance using only 2.31 M parameters and outperforms 12 state-of-the-art methods.

While spatial, temporal, and frequency domains are explored to enhance spectral efficiency (SE) in wireless communications, utilizing the polarization of electromagnetic (EM) waves also contributes to achieving this goal. Unlike conventional reconfigurable intelligent surface (RIS)-aided communications, polarization-transforming RIS (PTRIS) is first applied in this work to assist the line-of-sight (LoS) communication system. We introduce a novel framework for accurately modeling the direct and cascaded channels in the PTRIS-aided LoS communication system. This framework considers the spatial positions of the transmitter and receiver, the radiation patterns, antenna rotations, and the physical propagation mechanisms of EM waves based on antenna theory. The aperture field method is used to model the physical reflection of EM waves by PTRIS. Additionally, we aim to maximize SE by investigating four cases regarding the polarization-transforming capability of the PTRIS. Optimal closed-form solutions are derived for Case 1 and Case 2, while a best-effort approximation-based alternative optimization (BEA-AO) method is proposed for Case 3 to obtain sub-optimal solutions. Case 4 can be solved optimally with the barnch-and-cut algorithm within a reasonable computation time. Numerical results demonstrate that PTRIS can provide a robust and enhanced SE in LoS communications, even with arbitrary antenna rotations, compared to the scenarios without PTRIS.

Underwater images, often plagued by complex degradation, pose significant challenges for image enhancement. To address these challenges, the paper redefines underwater image enhancement as an image decomposition problem and proposes a deep invertible neural network (INN) that accurately predicts both the latent image and the degradation effects. Instead of using an explicit formation model to describe the degradation process, the INN adheres to the constraints of the image decomposition model, providing necessary regularization for model training, particularly in the absence of supervision on degradation effects. Taking into account the diverse scales of degradation factors, the INN is structured on a multi-scale basis to effectively manage the varied scales of degradation factors. Moreover, the INN incorporates several asymmetric design elements that are specifically optimized for the decomposition model and the unique physics of underwater imaging. Comprehensive experiments show that our approach provides significant performance improvement over existing methods.

Deep Video Quality Assessment (VQA) methods have shown impressive high-performance capabilities. Notably, no-reference (NR) VQA methods play a vital role in situations where obtaining reference videos is restricted or not feasible. Nevertheless, as more streaming videos are being created in ultra-high definition (e.g., 4K) to enrich viewers' experiences, the current deep VQA methods face unacceptable computational costs. Furthermore, the resizing, cropping, and local sampling techniques employed in these methods can compromise the details and content of original 4K videos, thereby negatively impacting quality assessment. In this paper, we propose a highly efficient and novel NR 4K VQA technology. Specifically, first, a novel data sampling and training strategy is proposed to tackle the problem of excessive resolution. This strategy allows the VQA Swin Transformer-based model to effectively train and make inferences using the full data of 4K videos on standard consumer-grade GPUs without compromising content or details. Second, a weighting and scoring scheme is developed to mimic the human subjective perception mode, which is achieved by considering the distinct impact of each sub-region within a 4K frame on the overall perception. Third, we incorporate the frequency domain information of video frames to better capture the details that affect video quality, consequently further improving the model's generalizability. To our knowledge, this is the first technology for the NR 4K VQA task. Thorough empirical studies demonstrate it not only significantly outperforms existing methods on a specialized 4K VQA dataset but also achieves state-of-the-art performance across multiple open-source NR video quality datasets.

Salient object detection (SOD) aims to identify the most prominent regions in images. However, the large model sizes, high computational costs, and slow inference speeds of existing RGB-D SOD models have hindered their deployment on real-world embedded devices. To address this issue, we propose a novel method named AirSOD, which is committed to lightweight RGB-D SOD. Specifically, we first design a hybrid feature extraction network, which includes the first three stages of MobileNetV2 and our Parallel Attention-Shift convolution (PAS) module. Using the novel PAS module enables capturing both long-range dependencies and local information to enhance the representation learning while significantly reducing the number of parameters and computational complexity. Secondly, we propose a Multi-level and Multi-modal feature Fusion (MMF) module to facilitate feature fusion, and a Multi-path enhancement for Feature Refinement (MFR) decoder for feature integration. The proposed method significantly reduces the model size by 63%, decreases the computational complexity by 43%, and improves the inference speed by 43% compared with the cutting-edge model (MobileSal). We test our AirSOD on six widely-used RGB-D SOD datasets. Extensive experimental results demonstrate that our method obtains satisfactory performance. The source codes will be made available.

Images taken on snowy days often suffer from severe negative visual effects caused by snowflakes. The task of removing snowflakes from a snowy image is known as image desnowing, which is challenging as image details are easily mistakenly treated and thus may be significantly lost during snowflake removal. Leveraging invertible neural networks (INNs), this paper presents a deep learning-based method for single image desnowing, which can remove snowflakes accurately while preserving image details well. Interpreting desnowing as an image decomposition problem, we propose an INN composed of two asymmetric interactive paths for predicting a latent image and a snowflake layer respectively. Such an INN is able to progressively refine the features of both latent images and snowflake layers for disentanglement, while retaining all information possibly relevant to latent image reconstruction. In addition, an attentive coupling layer supervised by snowflake masks is introduced to enhance feature dismantlement and a coupling-in-coupling structure is developed for further improvement. Extensive experiments show that, the proposed method outperforms existing ones on three benchmark datasets of synthetic and real-world images, and meanwhile it also shows advantages in terms of model size and computational efficiency.

In recent years, with the rise of intelligent hardware technology, the Internet of Things (IoT) has achieved rapid development and has been widely used in smart cities, smart power grids, industrial Internet, and other fields. The resulting security risks of IoT have attracted more and more attention. Being different from the traditional authentication that is based on MAC or security certificate, RF fingerprinting technology extracts fingerprints from the emissions of wireless transmitters. These fingerprints root in the hardware imperfection of the transmitting circuits and can be used for wireless device identification and authentication. RF fingerprinting can effectively enhance the security of the wireless network and has become a research focus in the field of IoT. However, how to learn and fuse multiple RF fingerprints is still an urgent problem to be solved. Inspired by the attention mechanism in the field of computer vision, an adaptive RF fingerprints fusion network (ARFNet) is proposed in this article, which is built on our dual attention convolution (DAConv) layer. This neural network can extract and adaptively fuse multiple RF fingerprints in a data-driven manner to obtain more discriminant features. In addition, a data augmentation method is also designed to effectively enhance the network’s robustness to channel variation and SNR variation. Extensive experimental results show that the proposed ARFNet combined with data augmentation can achieve 99.5% recognition accuracy on five USRP X310, and achieve 95.7% recognition accuracy on 56 ADS-B devices. Our source code has been released athttps://github.com/zhangweifeng1218/Adaptive_RF_Fingerprinting.

Although space-time adaptive processing (STAP) is recognized as the optimal clutter suppression way for synthetic aperture radar (SAR) in theory, the deficient of independent and identically distributed range samples in real scenario limits its application. The reduce-dimension STAP methods can decrease the demand for range samples, but the assumption of moving target-free is always unsatisfied. The direct data domain methods only use the data of the range cell under test (RCUT) to avoid the assumption, but they are conducive to interference suppression than clutter suppression and have huge computational burden. Thus, in this letter, a single range data-based STAP method is proposed not only exploring the space-time statistical properties of clutter to suppress it, but also operating solely on the RCUT without recourse to range samples. Theoretical analyses and simulation results verify the effectiveness of the proposed method.

Unknown motions will make the synthetic aperture radar (SAR) images of ground moving targets defocused. The target signal easily exhibits Doppler ambiguity due to the limitation of pulse repetition frequency, which leads to the focusing difficulty of moving targets. To address these issues, a fast approach for SAR imaging of ground moving target with Doppler ambiguity is proposed. In this method, the first-order and quadratic phase are initially estimated by using proposed operations based on 2-D scaled Fourier transform and improved range frequency cross correlation function, respectively. With the estimated parameters, the moving target is then focused in the range–azimuth time domain by the matched filtering. The presented approach is fast, because its realization procedure does not have any parameter-searching step and can be sped up by nonuniform fast Fourier transform. Moreover, the proposed approach can handle Doppler ambiguity (including Doppler center blur and spectrum ambiguity), blind speed sidelobe, and scaled frequency spectrum aliasing. Both spaceborne and airborne real data-processing results are presented to confirm the effectiveness of the proposed method.

This letter introduces a novel Fourier-based data augmentation strategy for visible-thermal person re-identification (VT-ReID). Different from some existing methods which are proposed from the perspective of network structure and loss functions, our method aims to fully consider the semantic information from the perspective of data preprocessing. The main hypothesis is that the phase component in the Fourier domain contains high-level semantic information and the amplitude component contains low-level modality awareness information. In order to make the model pay more attention to semantic information learning, we design a simple but effective Fourier-based semantic augmentation (FSA) module, which can be inserted seamlessly into any existing models. Extensive experiments on RegDB and SYSU-MM01 datasets have shown that our proposed method can improve the VT-ReID performance significantly and achieve state-of-the-art performance.

Synthetic aperture radar raw data simulation (SAR-RDS) is beneficial to the SAR system design, signal processing method verification, and radar parameter optimization. Most SAR-RDS methods are based on the flat terrain assumption. However, the fluctuant terrain in real scene will induce severe SAR beam occlusion effect and produce radar shadow, leading to incorrect RDS results. Thus, a dynamic elevation angle interpolation (DEAI) algorithm is proposed for SAR shadow judgment by considering the actual SAR working process. The key of the proposed DEAI algorithm is the 1-D EAI and shadow visualization update, which avoids the problem that the existing methods cannot judge the shadow of partial areas due to the insufficiently refined mesh grid or the mismatch of the judgment model. Moreover, an evaluation framework named as joint image and signal criteria (JISC) is proposed from the perspectives of SAR imaging and signal processing results to objectively evaluate the SAR-RDS results and solve the problem that the existing evaluation methods cannot be compatible with fluctuant terrain. Finally, the numerical experiment verified our theoretical analyses.

Owing to the highly maneuvering character involved in targets, the nonuniform 3-D rotation motions make the assumption that the image projection plane (IPP) is constant during coherent processing interval (CPI) invalid. In this work, an efficient approach is proposed in ISAR imaging for highly maneuvering targets with nonstationary IPP. First, to reasonably describe the mobility of a highly maneuvering motion target, the geometry and signal model with nonstationary IPP are established, where the high-order phase model is deduced to describe the 2-D spatial-variant phase errors. Second, based on the developed signal model, considering the cost function obtained via conventional image entropy with local extremum, the subarray averaging operation in conjunction with entropy is utilized to accelerate the global optimal convergence. Finally, the accurate 2-D spatial-variant phase errors compensation terms are generated to produce the well-focused ISAR images. Compared with existing methods, the main advantages of this work are: 1) the geometry and signal model of the target with nonstationary IPP are established; 2) the subarray averaging operation in conjunction with image entropy is utilized to accelerate the global optimal convergence; and 3) the high-order signal model is derived to present the 2-D spatial-variant phase errors. Several numerical experiments using simulated data and electromagnetic data are conducted to demonstrate the validity of the proposed algorithm and signal model.

Laryngeal cancer tumor (LCT) grading is a challenging task in P63 Immunohistochemical (IHC) histopathology images due to small differences between LCT levels in pathology images, the lack of precision in lesion regions of interest (LROIs) and the paucity of LCT pathology image samples. The key to solving the LCT grading problem is to transfer knowledge from other images and to identify more accurate LROIs, but the following problems occur: 1) transferring knowledge without a priori experience often causes negative transfer and creates a heavy workload due to the abundance of image types, and 2) convolutional neural networks (CNNs) constructing deep models by stacking cannot sufficiently identify LROIs, often deviate significantly from the LROIs focused on by experienced pathologists, and are prone to providing misleading second opinions. So we propose a novel fusion attention block network (FABNet) to address these problems. First, we propose a model transfer method based on clinical a priori experience and sample analysis (CPESA) that analyzes the transfer ability by integrating clinical a priori experience using indicators such as the relationship between the cancer onset location and morphology and the texture and staining degree of cell nuclei in histopathology images; our method further validates these indicators by the probability distribution of cancer image samples. Then, we propose a fusion attention block (FAB) structure, which can both provide an advanced non-uniform sparse representation of images and extract spatial relationship information between nuclei; consequently, the LROI can be more accurate and more relevant to pathologists. We conducted extensive experiments, compared with the best Baseline model, the classification accuracy is improved 25%, and It is demonstrated that FABNet performs better on different cancer pathology image datasets and outperforms other state of the art (SOTA) models.

The fifth generation New Radio vehicle-to-everything (5G NR-V2X) has higher requirements for delay and reliability, which brings challenges to the physical layer signal processing. Mobile edge computing (MEC) can provide larger capacity data storage and more efficient computation through localization of on-board unit (OBU) and next generation Node B (gNB) services. This paper combines the channel estimation of New Radio vehicle-to-infrastructure (NR-V2I) communication baseband signal processing with MEC and designs an intelligent channel estimation framework. In the MEC server, this paper proposes a channel estimation algorithm based on deep learning. This algorithm uses a one-dimensional convolutional neural network (1D CNN) to complete frequency-domain interpolation and conditional recurrent unit (CRU) for time-domain state prediction. Additional velocity coding vector and multipath coding vector track changes in the environment, and accurately train channel data in different mobile environments. System simulation and analysis show that the proposed algorithm improves the channel estimation accuracy, reduces the bit error rate, and enhances the robustness compared with the representative channel estimation algorithms for NR-V2I.

This letter presents a conceptually simple and effective dual-granularity triplet loss for visible-thermal person re-identification (VT-ReID). Generally, ReID models are always trained with the sample-based triplet loss and identification loss from the fine granularity level. Further, center-based loss could be introduced to encourage the intra-class compactness and inter-class discrimination from the coarse granularity level. Our proposed dual-granularity triplet loss well organizes the sample-based triplet loss and center-based triplet loss in a hierarchical fine to coarse granularity manner, just with some simple configurations of typical operations, such as pooling and batch normalization. Experiments on RegDB and SYSU-MM01 datasets show that with only the global features our dual-granularity triplet loss can improve the VT-ReID performance by a significant margin. It can be a strong VT-ReID baseline to boost future research with high quality.

This paper focuses on the visible-thermal cross-modality person re-identification (VT Re-ID) task, whose goal is to match person images between the daytime visible modality and the nighttime thermal modality. The two-stream network is usually adopted to address the cross-modality discrepancy, the most challenging problem for VT Re-ID, by learning the multi-modality person features. In this paper, we explore how many parameters a two-stream network should share, which is still not well investigated in the existing literature. By splitting the ResNet50 model to construct the modality-specific feature extraction network and modality-sharing feature embedding network, we experimentally demonstrate the effect of parameter sharing of two-stream network for VT Re-ID. Moreover, in the framework of part-level person feature learning, we propose the hetero-center triplet loss to relax the strict constraint of traditional triplet loss by replacing the comparison of theanchor to all the other samplesby theanchor center to all the other centers. With extremely simple means, the proposed method can significantly improve the VT Re-ID performance. The experimental results on two datasets show that our proposed method distinctly outperforms the state-of-the-art methods by large margins, especially on the RegDB dataset achieving superior performance, rank1/mAP/mINP 91.05%/83.28%/68.84%. It can be a new baseline for VT Re-ID, with a simple but effective strategy.

Representation-based target detectors for hyperspectral imagery (HSI) have recently aroused a lot of interests. However, existing methods ignore the dictionary structure and cannot guarantee an informative and discriminative representation of test pixels for target detection. To alleviate the problem, this letter proposes a novel sparse and dense hybrid representation-based target detector (SDRD). The proposed detector adopts the idea that the relationship between the background and the target sub-dictionaries is a collaborative competition. The structure of the dictionary is discovered and preserved by learning a sparse and dense hybrid representation for test pixel. Benefitting from this, a compact and discriminative representation can be obtained to better represent the test pixel for an improved detection performance. Experimental results on several HSI data sets verify the effectiveness of SDRD in comparison with several state-of-the-art methods.

Owing to the large range cell migration (RCM) and fast time-variant Doppler frequency modulation (DFM) generated by rapidly spinning targets, it is difficult to efficiently obtain well-focused inverse synthetic aperture radar (ISAR) images via conventional algorithms because of the multidimensional search requirement. Inspired by the inherent azimuth spatial invariance in strip-map synthetic aperture radar (SAR) imaging mode, an efficient range-Doppler domain ISAR imaging method for rapidly spinning targets is proposed in this article. First, echo signal is transformed into range-Doppler domain and its precise analytical expression is derived according to the principle of stationary phase (POSP). Second, the energy of scatterers distributed in different range cells is extracted along the rotating radius. By doing so, the energy is concentrated in the same range cell. After that, the high-order phase terms of the signal are compensated and the CLEAN technique is also applied to reduce the sidelobes of a strong scatterer. Finally, 3-D ISAR image of the spinning target is reconstructed by projecting the spatial parameters to 3-D cylindrical coordinates. Furthermore, in this article, the output signal-to-noise ratio (SNR) gain, anti-noise performance, the mismatched phase error, and the computational complexity analyses are also provided. Compared with existing approaches, the proposed method has advantages in the computational complexity and low SNR environment thanks to the only 1-D search and the coherent integration gain obtained. Both the theoretical derivations and the simulated results demonstrate the effectiveness of the proposed method.

Electronic tongue (E-Tongue), as a novel taste analysis tool, shows a promising perspective for taste recognition. In this paper, we constructed a voltammetric E-Tongue system and measured 13 different kinds of liquid samples, such as tea, wine, beverage, functional materials, etc. Owing to the noise of system and a variety of environmental conditions, the acquired E-Tongue data shows inseparable patterns. To this end, from the viewpoint of algorithm, we propose a local discriminant preservation projection (LDPP) model, an under-studied subspace learning algorithm, that concerns the local discrimination and neighborhood structure preservation. In contrast with other conventional subspace projection methods, LDPP has two merits. On one hand, with local discrimination it has a higher tolerance to abnormal data or outliers. On the other hand, it can project the data to a more separable space with local structure preservation. Further, support vector machine, extreme learning machine (ELM), and kernelized ELM (KELM) have been used as classifiers for taste recognition in E-Tongue. Experimental results demonstrate that the proposed E-Tongue is effective for multiple tastes recognition in both efficiency and effectiveness. Particularly, the proposed LDPP-based KELM classifier model achieves the best taste recognition performance of 98%. The developed benchmark data sets and codes will be released and downloaded in http://www.leizhang.tk/ tempcode.html.

To better interpret the inverse synthetic aperture radar (ISAR) imaging results, it is highly desirable to present them in the homogeneous range-cross-range domain, rather than the conventional range-Doppler (RD) domain. This process is referred to as cross-range scaling and the rotating angle velocity (RAV) of the moving target must be estimated first to achieve that goal. In this paper, an efficient cross-range scaling approach based on 2-D discrete wavelet transform (2D-DWT) and pseudopolar fast Fourier transform (PPFFT) is developed. To be exact, first, 2D-DWT is applied to two sequential ISAR images to obtain the dominant feature points based on the fact that the ISAR images are usually redundant for estimating RAV. By doing so, the data dimensional reduction and noise suppression are also realized. After that, second, via the efficient PPFFT, two sequential RD ISAR images are mapped into the pseudopolar coordinate to convert the rotational motion into the translational motion along the pseudo angle direction. Finally, to estimate the RAV, a new normalized correlation cost function is constructed and the Golden section algorithm is employed to efficiently find the optimal RAV. Compared with the conventional methods, the advantages of the proposed method are threefold: 1) the rotation center of a target is no longer required prior; 2) without the interpolation operation and the utilization of data dimensional reduction via 2D-DWT, the computational complexity of the proposed method is significantly reduced;and 3) the accurate RAV estimation is achieved in the case of low signal-to-noise ratio condition. The results from both the simulated and the measured data demonstrate that the proposed approach outperforms the state-of-the-art algorithms in terms of the estimation accuracy and computational complexity.

An innovative iterative process is proposed to acquire the dynamic process of multichannel slotted ALOHA (S-ALOHA). It reveals the direct relation between the number of contending devices that perform their jth random access (RA) attempt at the ith RA slot and the newly arrived devices before the ith RA slot. These results allow engineers to analytically derive the probability density function of RA delay of multichannel S-ALOHA, as well as its cumulative density function and average value. Under stable RA attempts assumption, simplified form of the above analysis is given, with which we prove the number of preamble transmissions follows truncated geometric distribution. Taking the two traffic models proposed for machine type communications as examples, numerical results are presented to verify the effectiveness of the proposed iterative process and the accuracy of its simplified form, and illustrate the delay characteristics of simplified long term evolution RA channel.

A hybrid multi-user detector for code division multiple access (CDMA) based on stochastic Hopfield neural network (SHNN) and reduced detector (RD) is presented in this paper. Its purpose is to give an insight to the problems of CDMA and to propose novel algorithms for multi-user detection (MUD). An investigation on the nature of the local minima of the optimal multi-user detector's (OMD) objective function leads to the development of an efficient algorithm that can reduce significantly the size of the OMD optimization problem and seems to have a superior performance. Because the RD is based on digital signal processing (DSP) and SHNN is based on discrete Hopfield neural network (DHNN), the hybrid detector is a digital processing scheme which can be realized with the development of DSP. The performance of the hybrid detector is evaluated via simulations and it is shown to exceed that of other sub optimal receivers at a much lower computational cost in CDMA transmission case.