Gamma belief network (GBN), widely viewed as deep probabilistic topic models, has demonstrated its potential for uncovering multi-layer interpretable latent representations from text corpora. Its notable performance in document modeling largely arises from the expressive nature of gamma-distributed latent variables, which naturally capture sparsity, nonnegativity, skewness, heavy-tailed pattens, and from their seamless extension to multi-layer hierarchical structures. However, existing GBN and its variations are constrained by linear generative model, thereby limiting their expressiveness and applicability. To address this limitation, we introduce Generalized Gamma Belief Network (Generalized GBN), which extends original linear generative model to a more expressive non-linear generative model. Since parameters of Generalized GBN no longer possess an analytic conditional posterior, we further propose an upward-downward Weibull inference network to approximate posterior distribution of latent variables. The parameters of both generative model and inference network are jointly trained within variational inference framework. In addition, we provide theoretical analyses that demonstrate the effectiveness of Generalized GBN in modeling data variability and achieving disentangled representations. The former benefit arises from its hierarchical latent-variable structure, while the latter stems from its inherent ability to model sparsity. Finally, we conduct comprehensive experiments on both expressivity and disentangled representation learning tasks to evaluate the performance of Generalized GBN against Gaussian variational autoencoders serving as strong baseline models.

Traditional radar target recognition methods typically model only single echoes, neglecting the crucial information that domain knowledge can provide for understanding data. In this paper, we propose a knowledge embedding fusion (KEF) method for enhanced high-resolution range profile (HRRP) recognition, which utilizes the target state descriptions available during radar detection. KEF leverages a language model (LM) to integrate textual knowledge with echo features for fusion recognition. It consists of three components: HRRP feature extraction, measurement-based knowledge construction, and knowledge embedding fusion module. First, we perform feature extraction on the HRRP to obtain echo tokens. Next, in the knowledge construction module, the measurement statuses are standardized to a natural language format, and the LM is utilized to extract semantic information, resulting in text tokens. Finally, in the knowledge embedding fusion module, a cross-attention HRRP-text fusion strategy is employed to facilitate interaction between echo tokens and textual tokens. We also design a combination of HRRP-text matching loss and fusion classification loss to guide model training. Experiments are conducted on a real measured dataset, and the results indicate that KEF effectively enhances recognition performance across multiple scenarios compared with approaches that only utilize echoes. • We develop a knowledge embedding fusion based HRRP recognition method. • It converts measurement information into textual knowledge and utilizes a language model for representation modeling. • It employs a cross-attention mechanism to fully integrate measurement knowledge with HRRP features. • It uses classification loss and HRRP-text matching loss for joint optimization. • It can improve the HRRP target recognition performance in various scenarios.

Multimodal remote sensing combines optical and synthetic aperture radar (SAR) imagery to improve perception, yet real deployments face spatially varying degradations (e.g., clouds, low light, sensor interference) that can corrupt fusion. To make robustness measurable, we introduce a controlled mixed-severity setting in which only the optical stream is synthetically cloud-degraded while SAR remains intact, providing a standardized testbed for evaluating multimodal detection under modality imbalance. We further present CAIR-Net, a reliability–aware information routing network that follows adenoise-then-fuseprinciple: a Local Reliability Modulation (LRM) module learns soft, spatial reliability maps to suppress degraded regionsbeforecross-modal interaction, and a Global Information Selection Mechanism (GISM) performs confidence-aware expert routing across optical, fused, and SAR experts. On the mixed-severity benchmark, CAIR-Net consistently outperforms strong unimodal and fusion baselines and exhibits a substantially smaller performance drop under severe clouds (only a 7.3% AP reduction versus drops exceeding 25% for representative alternatives). These results indicate that explicit reliability modeling and quality-guided routing provide a practical path toward robust multimodal detection when one modality is partially or nearly completely occluded.

Image restoration under adverse weather conditions is critical for real-world applications. However, existing approaches mainly suffer from two fundamental limitations,i) the impractical requirement of prior degradation knowledge for task-specific model selection andii) performance degradation when handling with in-the-wild corruptions. To address the above issues, in this paper, we propose a novelMeta-prior Aided Transformerrestoration framework, MePAT, to synergize dynamic feature modulation with optimal transport (OT) theory. Specifically, we first architect an efficient attention mechanism,rectified self-channel attention(RSCA) to capture long-range associations along the channel dimension. Then, to adaptively tackle different conditions, we design atask-shared prior learning network(TPLN) to generate content-adaptive weather embeddings and serve as feature modulators to direct a more flexible and robust restoration process. In addition to learn discriminative task features, we propose an weakly-supervised OT-driven contrastive loss to measure the discrepancy between different weather corruptions. During the inference process, through the shared TPLN, we derive image-oriented vectors for unseen corruptions and then perform image restoration. The superior experimental results on three synthetic benchmarks demonstrate the effectiveness of MePAT. We also conduct experiments on real-world applications to verify the generalization ability and robustness. The code and pre-trained models will be made available.

Concept Bottleneck Models (CBMs) try to make the decision-making process transparent by exploring an intermediate concept space between the input image and the output prediction. Existing CBMs just learn coarse-grained relations between the whole image and the concepts, less considering local image information, leading to two main drawbacks: i) they often produce spurious visual-concept relations, hence decreasing model reliability; and ii) though CBMs could explain the importance of every concept to the final prediction, it is still challenging to tell which visual region produces the prediction. To solve these problems, this paper proposes a Disentangled Optimal Transport CBM (DOT-CBM) framework to explore fine-grained visual-concept relations between local image patches and concepts. Specifically, we model the concept prediction process as a transportation problem between the patches and concepts, thereby achieving explicit fine-grained feature alignment. We also incorporate orthogonal projection losses within the modality to enhance local feature disentanglement. To further address the shortcut issues caused by statistical biases in the data, we utilize the visual saliency map and concept label statistics as transportation priors. Thus, DOT-CBM can visualize inversion heatmaps, provide more reliable concept predictions, and produce more accurate class predictions. Comprehensive experiments demonstrate that our proposed DOT-CBM achieves SOTA performance on several tasks, including image classification, local part detection and out-of-distribution generalization. Codes are available in supplementary material.

Concept-based models can map black-box representations to human-understandable concepts, which makes the decision-making process more transparent and then allows users to understand the reason behind predictions. However, domain-specific concepts often impact the final predictions, which subsequently undermine the model generalization capabilities, and prevent the model from being used in high-stake applications. In this paper, we propose a novel Language-guided Concept-Erasing (LanCE) framework. In particular, we empirically demonstrate that pre-trained vision-language models (VLMs) can approximate distinct visual domain shifts via domain descriptors while prompting large Language Models (LLMs) can easily simulate a wide range of descriptors of unseen visual domains. Then, we introduce a novel plug-in domain descriptor orthogonality (DDO) regularizer to mitigate the impact of these domain-specific concepts on the final predictions. Notably, the DDO regularizer is agnostic to the design of concept-based models and we integrate it into several prevailing models. Through evaluation of domain generalization on four standard benchmarks and three newly introduced benchmarks, we demonstrate that DDO can significantly improve the out-of-distribution (OOD) generalization over the previous state-of-the-art concept-based models. Our code is available at https://github.com/joeyz0z/LanCE.

Diffusion models have demonstrated outstanding performance in industrial anomaly detection. However, their iterative denoising nature results in slow inference speed, limiting their practicality for real-time industrial deployment. To address this challenge, we propose OmiAD, a one-step masked diffusion model for multi-class anomaly detection, derived from a well-designed multi-step Adaptive Masked Diffusion Model (AMDM) and compressed using Adversarial Score Distillation (ASD). OmiAD first introduces AMDM, equipped with an adaptive masking strategy that dynamically adjusts masking patterns based on noise levels and encourages the model to reconstruct anomalies as normal counterparts by leveraging broader context, to reduce the pixel-level shortcut reliance. Then, ASD is developed to compress the multi-step diffusion process into a single-step generator by score distillation and incorporating a shared-weight discriminator effectively reusing parameters while significantly improving both inference efficiency and detection performance. The effectiveness of OmiAD is validated on four diverse datasets, achieving state-of-the-art performance across seven metrics while delivering a remarkable inference speedup.

Many large-scale systems rely on high-quality deep representations (embeddings) to facilitate tasks like retrieval, search, and generative modeling. Matryoshka Representation Learning (MRL) recently emerged as a solution for adaptive embedding lengths, but it requires full model retraining and suffers from noticeable performance degradations at short lengths. In this paper, we show that sparse coding offers a compelling alternative for achieving adaptive representation with minimal overhead and higher fidelity. We propose Contrastive Sparse Representation (CSR), a method that specifies pre-trained embeddings into a high-dimensional but selectively activated feature space. By leveraging lightweight autoencoding and task-aware contrastive objectives, CSR preserves semantic quality while allowing flexible, cost-effective inference at different sparsity levels. Extensive experiments on image, text, and multimodal benchmarks demonstrate that CSR consistently outperforms MRL in terms of both accuracy and retrieval speed—often by large margins—while also cutting training time to a fraction of that required by MRL. Our results establish sparse coding as a powerful paradigm for adaptive representation learning in real-world applications where efficiency and fidelity are both paramount. Code is available at this URL.

The autonomous underwater vehicle (AUV) are experiencing widely deployment for the advantages of low cost and flexibility. However, the open light-of-sight communication links between clusters to clusters (C2C) in flying ad-hoc networks (FANETs) are vulnerable to malicious jamming in communication denied environments (CDEs), leading to the C2C communication with low-signal-to-interference-plus-noise ratio (SINR). Aiming at the collaborative search inefficiency resulting from unreliable C2C communication caused by low SINR in existing researches, we propose a distributed collaborative search planning method based on anti-jamming strategy by virtual multiple input-multiple output (VMIMO) in FANETs. First, we formulate the anti-jamming C2C collaborative communication model based on VMIMO to combat the malicious jamming utilizing the advantage of large-scale AUVs under limited resource. Then, we model the overall search objective function, considering search, connectivity and collision avoidance between AUV clusters. And we establish a distributed collaborative search optimization problem model based on distributed model predictive control (DMPC) framework, so that all AUV clusters can optimize the overall search objective by interacting with neighbor clusters. Finally, we propose a three-level collaborative optimization problem model with joint considering search, anti-jamming communication and network connectivity. To calculate the optimal solution, an alternating optimization method based on distributed stochastic algorithm is proposed to solve the three-level collaborative optimization problem model. Moreover, we conduct extensive simulations and the results show that the proposed method is efficient in terms of the anti-jamming ability and search efficiency compared with other state-of-the-art algorithms without cooperative strategy.

In the classical radar imaging framework, the echo signals can be well compressed and focused by matched filtering. Yet these methods suffered model mismatch in the non-idea scenarios, such as the active jamming, the motion errors. In these situations, the imaging results were defocused or blurred. To solve these problems, a new learnable SAR imaging method was proposed in this paper. First, a hierarchical U-shaped network ImagingNet was constructed to learn the imaging mechanism from the history data. The base model was formed by a training strategy to optimize the errors between the learning imaging result and the reference image. On this basis, a new teacher-student training strategy was developed to refine the base model, and form the advanced model accordingly. Different from the classical imaging framework, the proposed method could focus the echo signals in the ideal and non-idea scenarios. In addition, the proposed method could achieve the real-time imaging when deployed on the parallel computing platform. Multiple rounds of experiments were performed to verify the proposed method. The performance improvement of 0.209 and 0.502 for SSIM, 4.8 dB and 4.4 dB for PSNR were achieved in the active jamming and motion errors scenarios in comparison to the classical method.

SAR played an important role in target monitoring and identifying. Yet the imaging resolution and the spatial coverage was contradictory. A great many works were presented previously, yet little studies were devoted to complex-valued SAR image. In addition, the evaluation strategy customized to SAR image was unavailable. To address the problems, a new SAR image super-resolution method was proposed in this paper. Different from the preceding works, SAR super-resolution was achieved from the perspectives of metric-driven and task-driven, forming the base model (SARSR) and the advanced model (DetSR). The base model was composed of three phases, pre-imaging, feature refinement, and reconstruction. The periodogram spectral estimation technique was first used to achieve spatial alignment. The spectrogram was fed into a hybrid architecture to learn the high-level representations. The learned features were refined and reconstruction. On the basis, the prior knowledge transferred from target detection task was then used to promote the super-resolution performance, forming the advanced model. Finally, the evaluation system customized to SAR image was presented. A set of metrics composed of vision and imaging measurements were defined to assess the image quality initially. The task-driven evaluation was then presented. Multiple rounds of experiments were performed to verify the proposed method. The results proved that the metrics can be improved more than 10%, while the detection accuracy can be improved by 4%.

In the field of synthetic aperture radar automatic target recognition (SAR ATR), inherent distributional discrepancies between electromagnetic synthetic and measured SAR images pose significant challenges to the potential applications of the former. To bridge the gap, a novel unsupervised domain adaptation framework based on Multi-modal Feature fusion and global-local Joint Alignment (MFJA) is proposed in this article. The multi-modal feature fusion focuses on describing each target more comprehensively by leveraging both visual and scattering topological information. In the visual branch, the full-aperture image is decomposed into multiple sub-aperture images to explore the scattering variations of the target at different azimuths, facilitating a richer visual description. Meanwhile, both local scattering and spatial position information of keypoints are simultaneously integrated into the feature extraction in the scattering topological branch, promoting a more comprehensive scattering topological representation. Subsequently, a gated feature fusion module is developed to effectively fuse features derived from different modalities. The global-local joint alignment aims to align different domains with greater precision. Specifically, a power normalized weighted gradient reversal layer is proposed to guide the network to focus more on hard-to-align samples during global domain alignment, thus mitigating their interference with local domain alignment. While MFJA achieves satisfactory cross-domain recognition performance, its inference efficiency is somewhat constrained. Therefore, a domain-invariant cross-modal knowledge distillation (DCKD) algorithm with a tri-path collaborative alignment strategy is further developed to distill discriminative and domain-invariant knowledge from the multi-modal model into a compact visual model based on full-aperture images, thereby accelerating inference. Experiments conducted in three scenarios on the public Synthetic and Measured Paired Labeled Experiment (SAMPLE) dataset validate the effectiveness of both MFJA and DCKD.

In this paper, a joint beam selection and power allocation (JBSPA) scheme for multi-target tracking is proposed in a collocated MIMO (C-MIMO) radar network. The goal of this scheme is to achieve better resource utilization efficiency with a given resource budget. Under the condition of sufficient resources, the scheme minimizes the total resource consumption of the C-MIMO radar network. When the sensor resources are insufficient, the scheme maximizes the number of tracked targets that meet the tracking requirements. To evaluate the performance of multi-target tracking, we normalize and utilize the Bayesian Cramér-Rao lower bound (BCRLB) as the performance evaluation criterion. The JBSPA scheme is formulated as a non-convex optimization problem involving integer and continuous variables that are coupled. To address this problem, we propose a fast and effective three-step solution technique. Simulation results demonstrate that the proposed JBSPA scheme can save resources, significantly increase the target capacity, and improve the resource utilization efficiency of the C-MIMO radar network.

Recently, deep convolutional neural networks have been widely applied in target detection of synthetic aperture radar (SAR) images. However, the regular convolution kernel cannot effectively establish dependency between features of SAR image with geometric distortion. Meanwhile, SAR images contain a small number of vehicle targets, and the imbalance problem between foreground-background class is serious during training. To solve these problems, we propose a you only look once (YOLO) detector based on graph convolutional network (GCN) called GCN-YOLO. First, a multilayer GCN model called vision GNN (ViG) is used as feature extractor to model the local area and build long-term dependencies between features. In addition, a convolutional block attention module (CBAM) is embedded into the last layer to enhance semantic features. Then, we introduce the VariFocal loss (VFL) as confidence loss to relief the imbalance problem between positive and negative samples. The experimental results on the miniSAR data demonstrate the effectiveness of the proposed method.

With the development of jamming systems, the jammer is becoming much smarter and can change its strategy with radar, which forms a competitive game between the radar and jammer and poses a significant threat to the radar. In this paper, the anti-jamming game of frequency-agile (FA) radar is investigated, in which the jammer is capable of transmitting jamming signals with multiple frequency channels and different power allocation manners. To characterize the sequential interaction and partial observation of the anti-jamming game, an extensive-form game is introduced to model the confrontation between the radar and jammer. Subsequently, a novel multi-agent reinforcement learning (MARL) algorithm, termed NFSP-DDPG, is devised by combining neural fictitious self-play (NFSP) with deep deterministic policy gradient (DDPG) and modifying the supervised learning process of NFSP to solve Nash equilibrium (NE) strategies, which can handle the anti-jamming games featuring both discrete and continuous action spaces. Finally, simulation results show that the learning strategies acquired through the proposed method approximate NE and outperform the rule-based strategies. The signal-to-interference-pulse-noise ratio (SINR) improvements of radar taking NE strategy are 9.13dB, 9.12dB, and 9.15dB in the worst-case in the anti-jamming game with five frequencies compared to three rule-based strategies.

Large-magnitude errors, known as outliers, have a significant impact on the accuracy of elliptic positioning (EP) for a distributed MIMO radar. Most of the existing EP techniques are developed in the least squares sense, susceptible to outliers. In this letter, we propose to leverage a sparse argument to model the outliers and decompose the measurement noise into two components, namely, inliers and outliers. By doing so, the robust EP is formulated as a$\ell _{0}$-norm minimization problem with non-convex constraints. By relaxing the objective$\ell _{0}$-norm with the convex$\ell _{1}$-norm, an alternating direction method of multipliers based algorithm is derived to efficiently solve the relaxed optimization problem. Numerical results illustrate the improved performance of the proposed method compared to the existing methods.

The acquisition of a well-focused and high-resolution inverse synthetic aperture radar (ISAR) image of a ship target is crucial for accurate target classification and recognition. In practical scenarios, ship targets exhibit complex maneuverability, and radar platforms demonstrate nonstationary behavior, which poses a serious challenge to conventional ISAR imaging algorithms. To address this problem, this article proposes a high-success-rate and fast ISAR imaging algorithm of nonstationary moving platform-attitude rapidly changing ship target (NSMP-ARCST) with Dempster-Shafer (DS) evidence and minimum entropy theory. The proposed algorithm employs multiple metrics to evaluate the imaging results and leverages the DS evidence theory to fuse these metrics for optimal imaging time interval (OITI) selection, aiming to enhance the success rate and robustness of ISAR imaging. Furthermore, this article introduces an improved fixed-point iterative minimum entropy phase-adjustment (IFPI-MEPA) method to optimize the ISAR imaging quality and computational speed under low signal-to-noise ratio (SNR) conditions, which contributes to an increased success rate of OITI selection and reduced computational complexity, thereby endowing it with substantial practical applicability. Experimental results using both simulated and real measured data illustrate the effectiveness and robustness of the proposed algorithm.

The high orbit height leads to the ultralow signal-to-noise ratio (ULSNR) of geosynchronous (GEO) targets, so it is necessary to increase the coherent processing interval (CPI) to improve the coherent accumulation gain. The ultra-long CPI (ULCPI) causes wide rotational angles and complex signal modulation, which poses a serious challenge to conventional inverse synthetic aperture radar (ISAR) imaging algorithms. Moreover, the high orbit height and high-speed motion of GEO targets make the “stop-and-go” model no longer applicable, and intra-pulse motion errors must be taken into consideration. To address the problems, this article proposes a high-resolution ISAR imaging algorithm with joint intra-pulse and inter-pulse high-order motion compensation (JIPHOMC) and sub-aperture images fusion for GEO targets at ULCPI. In this technique, ULCPI is divided into several sub-apertures. Then, with respect to the motion compensation in sub-apertures, we innovatively propose a JIPHOMC algorithm with particle swarm optimization (PSO), which can correct the complex spatial-time-variant (STV) motion errors caused by wide rotational angles and high-speed motion, and eliminate the 2-D high-order coupling. By utilizing image overall constraints, JIPHOMC avoids signal source decomposition processing, thereby boosting algorithm efficiency and robustness. For the sub-aperture images obtained from different perspectives, we develop a sub-aperture image fusion algorithm (SAIF) based on non-negative matrix factorization (NMF) of structured weighted sparse enhancement, which can ensure the integrity of the target structure and enhance the image signal-to-noise ratio (SNR), so as to achieve high-resolution ISAR imaging of GEO targets at ULCPI. Extensive experiments based on both scattering point simulation data and electromagnetic calculation data corroborate that the proposed algorithm outperforms traditional ISAR imaging methods for GEO targets at ULCPI.

Recently, metric-based meta-learning has been widely adopted to solve few-shot synthetic aperture radar (SAR) target classification, where the global features are employed to measure the similarity of each sample and the class prototypes. However, due to the high inter-class similarity of SAR images, the global features may suppress the detailed information in local features beneficial for SAR target classification. Besides, the prototypes obtained from a few SAR support samples at sparse azimuth angles tend to be biased. To tackle the above problems, we first integrate a local feature classification module into meta-learning and propose a multi-scale local classification network (MLC-Net) to enhance the target’s critical local detail features. Then, the feature generation and calibration network (FGC-Net) is proposed to generate SAR support features at the full range of azimuth angles to compensate for the real support features extracted by MLC-Net. Specifically, FGC-Net consists of a feature generative adversarial network (FGAN) and an adaptive feature calibration module (AFCM). FGAN is designed to generate support features for each class under the full range of azimuth angles. AFCM is proposed to calibrate the generated support features by adaptively re-weighting the generated and real support features of the same class. Finally, we devise the prototype mean square error (PMSE) loss and the central constraint (CC) loss to further narrow the distribution margin between the generated and real support features. FGC-Net and MLC-Net are unified for end-to-end training, resulting in consistent performance gains in feature generation and few-shot classification. Extensive experiments on the moving and stationary target acquisition and recognition (MSTAR) benchmark dataset under different few-shot settings demonstrate the effectiveness of our proposed method.

Image registration is the basis for joint utilization of multisource scene information. However, accurate automatic registration of multisource remote sensing images remains a challenging task, especially for optical and synthetic aperture radar (SAR) images. Due to the large geometric and intensity differences between images, many algorithms often fail to accurately register or even mismatch. In this article, we propose a novel coarse-to-fine method with stable high accuracy, which mainly consists of three steps. First, extract consistent features for robust coarse registration. Considering the differences in local gradient magnitudes between optical and SAR images, image intensity preprocessing is performed. Meanwhile, instead of the simple calculations via horizontal and vertical gradient operators, multidirectional gradient operators are defined in two scale spaces to obtain consistent local gradient information. Via multidirectional consistent gradients, highly repeatable keypoints are detected. On the multidirectional gradient maps, support regions with multiple scales are utilized to construct multiscale and multidirectional consistent cross-modal feature descriptors. Second, match the extracted features. Aiming to obtain a reliable alignment in coarse registration step, novel strategies are implemented for the first matching of a cascaded feature matching method. Sufficiently reliable initial matches are established via a new two-way matching strategy, and obvious outliers are removed by exploring the consistencies of both spatial scale and local neighborhood elements of the correct matches. Third, form distinctive pixelwise feature representations for accurate fine registration. In order to increase the distinctiveness of features to distinguish adjacent pixels, a new filter bank based on small receptive fields is defined. Fine features are constructed, appearing as thinner structures at the edges. Meanwhile, through multiscale and multidirectional convolutions, sufficient neighborhood information is mined. Therefore, more precise correspondences can be found to fine-tune the roughly corrected image pair. Overall, a combination method is proposed with a feature-based method and an area-based method for coarse registration and fine registration, respectively. On simulated and real image pairs, the above three steps and the two-stage framework are verified. Experimental results show that the proposed optical-to-SAR image registration method based on the designed multiscale, multidirectional consistent feature and multiscale, multidirectional fine feature (M2F2M) is superior to the current representative feature-based and area-based methods in robustness and accuracy.

Target search, localization and tracking by unmanned aerial vehicle (UAV) swarms have attracted much attention recently in research and industrial applications. Consequently, there is a high demand for effective mission planning methods for UAV swarms in air transportation. However, it’s difficult to obtain the real-time mission planning in the uncertain dynamic environment. In response to tackle the problem, we model the mission planning problem in the uncertain dynamic environment as a dynamic multi-constraint and multi-objective optimization problems (DMCMOP) and propose the dynamic constrained two-archive evolutionary algorithm (DCTAEA) to realize the efficient mission planning. The proposed method can reconstruct the convergence archive (CA) and the diversity archive (DA) adaptively, and introduce the dynamic self-adaptive penalty mechanism into the CA updating, DA updating and the mating selection, which utilizes valuable infeasible solutions and promote population convergence. Consequently, the proposed algorithm can balance convergence, diversity and feasibility simultaneously. Comprehensive experiments on the real scenes and benchmark problem demonstrate that, compared with state-of-the-art algorithms, the proposed algorithm has the superiority and effectiveness.

Due to the complexity of real-world applications, open set recognition is often more practical than closed set recognition. Compared with closed set recognition, open set recognition needs not only to recognize known classes but also to identify unknown classes. Different from most of the current methods, we proposed three novel frameworks with kinetic pattern to address the open set recognition problems, and they are kinetic prototype framework (KPF), adversarial KPF (AKPF), and an upgraded version of the AKPF, AKPF++. First, KPF introduces a novel kinetic margin constraint radius, which can improve the compactness of the known features to increase the robustness for the unknowns. Based on KPF, AKPF can generate adversarial samples and add these samples into the training phase, which can improve the performance with the adversarial motion of the margin constraint radius. Compared with AKPF, AKPF++ further improves the performance by adding more generated data into the training phase. Extensive experimental results on various benchmark datasets indicate that the proposed frameworks with kinetic pattern are superior to other existing approaches and achieve the state-of-the-art performance.

The performance of the polarimetric notch filter (PNF) algorithm based on the orthogonal projection technique has been widely recognized in the field of polarimetric synthetic aperture radar (PolSAR) ship detection. However, for detection problems in non-orthogonal target subspace and clutter subspace, using the orthogonal projection technique results in detection performance loss. In this letter, we propose a new ship target detector based on the oblique projection technique to deal with the condition that the target subspace and the clutter subspace are non-orthogonal. Since most sea clutter is inconspicuous in low-to-medium sea conditions, it is time-consuming to apply a fine target detector over the entire scene. Therefore, this letter proposes a keypoint detector based on the noncircularity and polarimetric scattering information to achieve fast localization of potential target regions. In general, we propose a novel two-step ship detection method for PolSAR data. The first step is to use the keypoint detector for potential target region localization, and the second step is to use the oblique projection detector to perform fine target detection on the potential target regions. Experiments on the synthetic and measured PolSAR data demonstrate that the proposed method is effective for ship detection.

The multipath in very high-frequency (VHF) radar will reduce the direction-of-arrival (DOA) estimation accuracy for the low-angle target following. Since the target and multipath signals have strong coherence in the spatial domain, it is challenging to distinguish them accurately. Especially in some terrain complex regions, various scattering media cause the multipath echo to exhibit multichannel and nonuniform energy distribution features. Therefore, the single-multipath DOA estimation methods will mismatch the actual signal, resulting in inaccurate DOA estimation. The current letter presents a new DOA estimation approach using a multiple scattering center model in complex terrain to solve the mentioned issue. In the presented method, multiple multipath is analogous to a single one based on the multipath signal’s spatial coherence. The equivalent model represents the DOA estimation problem with a parametric optimization problem, and accurate estimation results can be attained using the Newton optimization approach. Compared with the conventional sparse reconstruction method, the presented method is not restricted by the restricted isometry property (RIP) and employs a robust parameter estimation method to solve the angle super-resolution problem under the low signal-to-noise ratio (SNR). The experimental results reflect that the presented model and approach can efficiently promote the DOA estimation precision and calculational performance compared to conventional DOA estimation approaches.

The recent years had witnessed a resurgence on neural network. Many hidden layers were stacked hierarchically to learn the high-level representations. Great performances were achieved by the learned representations. However, this kind of learning models were highly dependent on large amounts of signals with label information. In the realistic scenarios, it is very difficult and costly to collect the modulated signals with label information. Given small training samples, the fitting power of deep models were limited. To solve the problems, a new family of signal augmentation strategies, segment-wise generation and signal-wise generation are proposed. The former builds new signal by tuning a single signal, while the latter combines several different modulated signals together to produce new signal. Four kinds of techniques, segment shift in cyclic, segment correlation in random, pairwise signals combination, and multiple signals concatenation are presented. The aim is to simulate the unforeseen disturbances during signal sampling. The recognition performance under the realistic scenarios can be then improved. Multiple comparative studies were performed. The results demonstrated the effectiveness of proposed strategy in comparisons to the classical methods, as well as the deep learning algorithms.

This article considers the 3-D collaborative trajectory optimization (CTO) of multiple unmanned aerial vehicles to improve multitarget tracking performance with an asynchronous angle of arrival measurements. The predicted conditional Cramér–Rao lower bound is adopted as a performance measure to predict and subsequently control tracking error online. Then, the CTO problem is cast as a time-varying nonconvex problem subjected to constraints arising from dynamic and security (height, collision, and obstacle/target/threat avoidance). Finally, a comprehensive solution method (CSM) is presented to tackle the resulting problem, according to its unique structures. Specifically, if all security constraints are inactive, the CTO can be simplified as a nonconvex problem with convex dynamic constraints, which can be solved by the nonmonotone spectral projected gradient (NSPG) method. Oppositely, an alternating direction penalty method (ADPM) is presented to solve the CTO problem with some positive security constraints. The ADPM introduces auxiliary vectors to decouple the complex constraints and separates the CTO into several subproblems and tackles them alternately, while locally adjusting the penalty factor at each iteration. We show the subproblem w.r.t. the position vector is nonconvex but with convex constraints, which can be efficiently solved by the NSPG method. The subproblems w.r.t. the auxiliary vectors are separable and have closed-form solutions. Simulation results demonstrate that the CSM outperforms the unoptimized method in terms of tracking performance. Besides, the CSM achieves the near-optimal performance provided by the genetic algorithm with much lower computational complexity.

Target recognition via deep learning has achieved great performances in the preceding works. Yet this family of method are dependent on large amounts of training data with label information. For radar sensors, it is difficult to collect labeled data in practical due to the absent imaging truth. The commonly used solution is data augmentation. However, seldom studies are devoted to complex-valued radar images. In this paper, a new radar image generation method via target re-imaging is proposed. The original image is first cast into the frequency-aspect domain. The one axis represents the transmitted frequency, while the other presents the synthetic azimuth. The inverse operations of zero-padding and windowing are then applied on the transformed data. Two kinds of imaging techniques, the intra-sample re-imaging and the inter-sample re-imaging are presented to generate new radar images. The generated images are finally used to improve the learning effectiveness of deep models. Multiple comparative studies are performed to demonstrate the advantages of the proposed method.

In this article, we propose an efficient direction of departure (DOD) and direction of arrival (DOA) estimation method for bistatic multiple-input multiple-output (MIMO) radar with faulty arrays. A third-order tensor model is built, and the measurement 3-D structure can be better utilized than the traditional matrix model. Subsequently, the atomic norm minimization (ANM) technique is used to further improve the angle estimation performance. Furthermore, we found in the research process that when the faulty arrays still maintain the symmetry property, the measurement tensor can be converted to the real-valued domain by the forward–backward averaging technique and the unitary transform technique. The new algorithm we proposed exploits the multidimensional structure of the signal without estimating the signal subspace. Comparing with traditional matrix completion (MC) methods, it has a better performance in terms of robustness and resolving correlated targets. Also, the algorithm proposed in this article does not require angle pairing. Simulation results verify the effectiveness of the proposed algorithm.

Coherent LiDAR, which is based on triangular frequency modulated continuous wave (T-FMCW), is a promising technology for high-resolution imaging, because it can achieve long-range and high-precision measurements using low power. However, owing to the short wavelength of laser, the system is extremely sensitive to time-varying vibration. Most vibration compensation methods perform well in the scenes that contain strong scatterers using enough data. However, they lose validity in scenes without strong scatterers and sufficient observation time. To address this problem, we develop a second-order vibration error model for coherent LiDAR that divides the Doppler frequency shift caused by time-varying vibrations into quadratic and primary errors. Based on the vibration error model, we propose a two-stage time-varying vibration compensation method using the adaptive differential evolution (ADE) algorithm. In the first stage, we iteratively determine the optimal quadratic vibration coefficients that minimize the spectral entropy of the compensated dechirp signals using the ADE algorithm. Subsequently, we compensate for the quadratic errors using the estimated quadratic vibration coefficients. In the second stage, we use the frequency spectrum cross-correlation method to estimate the primary vibration coefficients and compensate for the primary errors. We perform experiments in the scene containing a point target and the scenes containing multi-component targets and distributed targets without strong scatterers. The results indicate that the proposed method can effectively compensate for time-varying vibration errors using only one-period T-FMCW data in different scenes. Further, the results of the proposed method are more accurate and stable than the Doppler frequency shift method.

Radar automatic target recognition (RATR) based on high resolution range profile (HRRP) has attracted more attention in recent years. In fact, the actual application environment of RATR is open set environment rather than closed set environment. However, previous works mainly focus on closed set recognition, which classifies the known classes by dividing hyper-planes in the feature space, and it will cause classification errors in open set environment. Therefore, open set recognition is proposed to solve this problem, which needs to determine a closed classification boundary for the identification of the known and unknown targets simultaneously. To accomplish this purpose, this paper proposes and proves the extreme value boundary theorem, which demonstrates that the maximum distance from the known features to the cluster center follows the generalized extreme value distribution. According to the proposed theorem, the closed classification boundary of the cluster is easily determined to distinguish between the known and unknown classes. Finally, extensive experiments on measured HRRP data verify the validity of the proposed theorem and the effectiveness of the proposed method.

In recent years, how to accurately identify targets from measured synthetic aperture radar (SAR) images according to electromagnetic synthetic SAR images is attracting more and more research interest. Most existing algorithms only focus on decreasing the domain differences in visual representations, while the special characteristics of SAR images are not explored enough. Besides, these algorithms tend to align only the overall distribution of synthetic and measured images, while domain shifts between subclasses are ignored. To solve these problems, a novel unsupervised domain adaptation framework named visual and scattering topological feature fusion and alignment network (VSFA) is proposed in this article. First, considering that visual features are crucial for recognition, image reconstruction is introduced to enhance the generalization of visual features. Second, by analyzing the imaging mechanism of SAR, we explore the differences in scattering topologies between synthetic and measured images for the first time. To measure the differences quantitatively, we model the non-Euclidean scattering topology of the target as graph data and introduce graph neural networks (GNNs) to extract scattering topological features. Moreover, in order to describe the scattering topology of the target more comprehensively, we introduce two different but complementary scattering topological point extraction algorithms and achieve their fusion at the feature level by GNN for the first time. Finally, a simple but effective two-stage domain adaptation loss is proposed to constrain the network to align the distribution of synthetic and measured images class by class. Benefiting from the simultaneous reduction of distribution differences in visual space and scattering topological space, the proposed method achieves 99.15% and 98.18% recognition accuracies in two typical experiment scenarios of the Synthetic and Measured Paired Labeled Experiment (SAMPLE) dataset, demonstrating its effectiveness.

Multiple unmanned aerial vehicle (Multi-UAV) cooperative area search is an important and effective means of intelligence acquisition and disaster rescue. Search path planning is a critical factor to improve multi-UAV search performance. Aiming at the search inefficiency resulting from insufficient cooperation between UAVs in existing researches, we present a novel distributed real-time search path planning method based on distributed model predictive control (DMPC) framework. Firstly, we formulate the overall search objective function in finite time domain, considering not only repeated searches, but also maintenance of connectivity and collision avoidance between UAVs. Secondly, we decompose the overall search objective function to establish a distributed constrained optimization problem (DCOP) model, so that all UAVs optimize the overall search objective by interacting with neighbors. Thirdly, aiming at the problem of falling into the local optima in existing algorithms, distributed stochastic algorithm based on enhanced genetic algorithm (DSA-EGA) is proposed to solve the established DCOP model. We design a point crossover operator and introduce anytime local search (ALS) framework that stores the global optimal solution explored. Finally, the simulation results of different benchmark problems demonstrate that the proposed DSA-EGA outperforms other state-of-the-art algorithms in terms of the quality of solution. The simulation results of cooperative area search problems illustrate that the established DCOP model improves the search efficiency by 7.7%, and DSA-EGA improves the search efficiency by 4.3% at least. In addition, we also verify that our method has high scalability.

The motion of sensors during the measurement period, if not accounted for, can degrade significantly the localization accuracy. This paper investigates the sensor motion effect for the positioning of an object, using angles of arrival only or together with time of arrival measurements. The biases from the AOA and TOA data models when ignoring the motion effect are examined. Positioning algorithms for AOA localization and mixed AOA-TOA localization that account for the motion effect are developed by the pseudo-linear formulation. The algorithms derived include the computationally attractive closed-form estimators and the noise resilient semidefinite programming solutions. The bias coming from the pseudo-linear formulation is analyzed in detail, and it can be subtracted from the closed-form solution to obtain a bias-suppressed estimate. Simulation validates the effectiveness of the proposed solutions in achieving the CRLB performance under Gaussian noise before the thresholding effect occurs.

This work addresses the source localization and association problem in a multipath propagation environment. By focusing on the limitation of the prior information in practical applications, we propose a target localization and association method based on iterative optimization with semi-unitary constraint and eigen-decomposition techniques. In contrast to the previous works, the proposed method can localize spatial sources and associate the incident paths to each source without prior knowledge pertaining to the propagation environment. Moreover, the proposed approach can be applied to an arbitrary array geometry without reducing the effective array aperture. Both simulations and real data experiments validate the effectiveness and robustness of the proposed method.

The huge data collected every minute present an urgent need for image interpretation automatically. Many studies were performed previously, yet the problem was still far from being solved. For the task of detection and recognition, the preceding works usually paid more attentions to feature extraction on a single image. The azimuth information were directly ignored. This is because SAR images of target in many azimuths were absent in practical. It is therefore difficult to achieve image interpretation from the perspective of azimuth. To solve the problem, this paper develop a new recurrent architecture. We regard the images of target in the consecutive azimuths as the samples of sequence signal at the specific times. A new recurrent neural network composed of some convolutional cells is then developed. It inputs the samples prior to the current azimuth, and outputs the prediction in the next azimuth. Multiple studies are performed to demonstrate the effectiveness of proposed strategy.

With the increasing requirements for target classification capabilities in an open scene, various open set recognition algorithms emerge in an endless stream. It becomes urgent to evaluate the recognition performance. However, very few studies have examined this crucial and challenging problem previously. To fill the gap, a new evaluation strategy is proposed in this paper. First, two measurements, Openness and openness based on the number of instances per unknown class are presented to quantify the degree of open scenario. The recognition performance is then evaluated by this unified evaluation criteria across several different perspectives, the number of classes, the number of instances, and the synthesis difficulty, on which the ability of algorithms to classify the known classes and identify the unknown classes can be analyzed quantitatively. Multiple experiments are performed on the real dataset. The experimental results demonstrate the effectiveness of the proposed evaluation strategy.

Ship classification from synthetic aperture radar (SAR) images tends to be a hotspot in the remote sensing community. Currently, more efforts have been made to the single-polarization (single-pol) SAR ship classification with limited performance. This letter proposes to explore the dual-polarization (dual-pol) SAR images for better ship classification. To be specific, a novel group bilinear convolutional neural network (GBCNN) model is developed to deeply extract discriminative second-order representations of ship targets from the pairwise VH and VV polarization SAR images. Particularly, the deep bilinear features are efficiently acquired by performing the bilinear pooling on sub-groups of deep feature maps derived, respectively, from the single-pol SAR images (self-bilinear pooling) and dual-pol SAR images (cross-bilinear pooling). To fully explore the polarization information, the multi-polarization fusion loss (MPFL) is constructed to train the proposed model for superior SAR ship representation learning. By extensive experiments, the proposed method can achieve an overall accuracy of 88.80% and 66.90% on the 3- and 5-category dual-pol OpenSARShip data sets, which outperform the state-of-the-art methods by at least 2.00% and 2.37%, respectively.

The$L$-type double baseline configuration (three antennas) radar system is commonly adopted to obtain 3-D images in the traditional interferometric inverse synthetic aperture radar (InISAR) imaging, making high demands on the complexity of hardware design and signal processing. In this letter, a novel InISAR imaging framework based on single baseline configuration (SBC) (two antennas) for maneuvering targets is proposed, which can acquire the range and azimuth coordinates of the targets through transmitting wideband signals and azimuth scaling. To address the problems of error transmission and insufficient robustness in the traditional cascaded motion compensation method, a joint motion compensation and azimuth scaling (JMCAS) algorithm is developed. By maximizing the image contrast (IC), this method can perform the optimal parameter estimation of translational and rotational motion of maneuvering targets, so as to simultaneously achieve the fine motion compensation and azimuth scaling. In addition, a non-coherent fusion image registration (NCFIR) algorithm is presented to achieve the image registration between the two antennas in a vertical direction. On this basis, the height coordinates of the targets can be obtained by means of interferometric processing. Extensive experimental results from both simulated and real data corroborate that the proposed algorithm can achieve high-precision 3-D imaging of maneuvering targets with low hardware complexity at a low cost.

Recently, deep neural network has achieved enormous success on target detection but how to better apply these methods to synthetic aperture radar (SAR) images is still a challenge due to the following two reasons: the characteristics of SAR images are not utilized sufficiently; SAR images contain a smaller number of targets, and thus detectors face more serious foreground-background class imbalance problem during training. In this paper, we propose a constant false alarm rate (CFAR) guided dual stream single shot multibox detector to solve the above problems. Firstly, based on the proposed dual sub-networks and the interactive channel-spatial attention fusion (ICSAF) module, we effectively utilize the strong scattering characteristic of SAR target to improve the feature extraction ability. Then we integrate the CFAR indicative map into focal loss to alleviate the foreground-background class imbalance problem. In addition, a non-maximum suppression method based on area-ratio (AR-NMS) is proposed for problem caused by dividing large scene SAR image into blocks. The whole framework can realize end-to-end joint training, and the experimental results on miniSAR real data demonstrate the effectiveness of the proposed method.

The open set recognition model can identify the known and unknown samples simultaneously. In the radar automatic target recognition application, open set recognition meets better practical requirements than closed set recognition. Theoretically, an overlap usually exists between the known and unknown features, which makes it difficult for the model to identify unknown samples. Therefore, we explore the distribution of the known and unknown features, and find that the unknown features are usually smaller and closer to the center region in the feature space than the known features. Based on this phenomenon, we propose a novel loss function that improves the open set recognition performance by controlling the known features distributed to the surrounding area of the feature space. In addition, extensive experiments are carried out on measured HRRP data. Thus, we verify the effectiveness of the proposed method.

Although several algorithms have achieved automatic registration on optical and synthetic aperture radar (SAR) images, it is still a challenge to establish enough reliable correspondences between such images due to their different imaging mechanisms. For this purpose, we propose a robust point-feature-based registration method. Considering the inherent properties of optical image and SAR image, two different gradient operators are utilized to construct scale spaces and extract features. A new gradient operator is defined for SAR image, yielding a more consistent gradient with optical image gradient calculated by the multiscale Sobel operator. Then, a novel cascaded matching method called cascaded sample consensus (CSC) is put forward to increase the number of correct correspondences. In the first matching, a simple but effective scale constraint strategy is used to remove outliers for a robust initial transformation model. Considering the spatial location relationship in each correct matching pair, the second matching constructs precise search spaces of the best matching points for more correspondences. Experimental results finally verify the robustness and accuracy of the proposed algorithm.

Due to the difficulties of obtaining sufficient real synthetic aperture radar (SAR) images, introducing simulated images can effectively enrich the training dataset in SAR target recognition. This letter explores how to accurately identify the targets in real SAR images by only using simulated training images. The key challenge is that there are distribution differences between the simulated (training) and real (test) data, which limits the performance of recognition methods. To solve this problem, a hierarchical recognition method is proposed. A well-trained convolutional neural network (CNN) is first utilized to pre-classify all the test images. Then, the focus of our proposed method is to find the hard test samples that are easy to be misclassified according to the CNN classification confidence and re-classify them. In fact, the distribution of these samples is relatively inconsistent with the distribution of the training data. Thus, we propose a multi-similarity fusion (MSF) classifier to re-classify them by comprehensively measuring the correlation between the hard samples and the training images through five similarity measures. During the fusion process, the bagging ensemble technique is used and the similarity measures are sampled to generate different subsets to enhance the diversity of sub-classifiers, thus the performance is improved. A large number of experiments finally verify the robustness and accuracy of the proposed method.

In this article, considering the supervised dimensionality reduction, we first propose a model, called infinite Bayesian max-margin linear discriminant projection (iMMLDP), by assembling a set of local regions, where we make use of Bayesian nonparametric priors to handle the model selection problem, for example, the underlying number of local regions. In each local region, our model jointly learns a discriminative subspace and the corresponding classifier. Under this framework, iMMLDP combines dimensionality reduction, clustering, and classification in a principled way. Moreover, to deal with more complex data, for example, a local nonlinear separable structure, we extend the linear projection to a nonlinear case based on the kernel trick and develop an infinite kernel max-margin discriminant projection (iKMMDP) model. Thanks to the conjugate property, the parameters in these two models can be inferred efficiently via the Gibbs sampler. Finally, we implement our models on synthesized and real-world data, including multimodally distributed datasets and measured radar image data, to validate their efficiency and effectiveness.

Computational microwave imaging (CMI) based on the frequency diversity metasurface apertures (FDMAs) is an emerging technology and has attracted wide attention. FDMA based CMI (FDMA-CMI) can be considered as microwave compressive sensing imaging with the frequency diversity pattern of the FDMA being the sensing matrix and solved by sparse signal reconstruction algorithms. However, the imaging quality is affected by the sensing matrix error and off-grid error seriously. In this paper, we propose a novel algorithm for FDMA-CMI, referred to as OGSISBL, by taking both the off-grid error and sensing matrix error into account. Firstly, we establish the measurement model with both the off-grid error and sensing matrix error. Specifically, the off-grid error is represented as a set of parameters to be estimated in the measurement model and the sensing matrix error is represented as the amplitude-phase drift of the transceiver channels of the imaging system due to the principle of the FDMA. Then, under the framework of the sparse Bayesian learning, a robust imaging algorithm OGSISBL is developed via the variational Bayesian expectation maximization (VBEM), which can not only recover the amplitude and position of the return of the scattered, but also simultaneously calibrate the amplitude-phase drift of the transceiver channels and the off-grid error. The performance of the proposed algorithm is evaluated by both the simulation data and the measured data collected by the self-designed experimental FDMA-CMI system, and the results validate the effectiveness and robustness of the proposed method.

In this article, a composed resource optimization (CRO) scheme is developed for an active and passive radar network engaged in multiple target tracking (MTT). The motivation of the CRO scheme is to collaboratively optimize the transmit resources of active radars, as well as the receiving processing resources of passive radars, to improve the overall MTT performance. We utilize the predicted conditional Cramér–Rao lower bound to evaluate the impact of allocation strategies on tracking performance and formulate the CRO as a mixed-integer nonlinear program problem since the adaptable parameters w.r.t. the target selection process are in binary form. To solve the problem, we propose an alternating direction method of multiplier-based algorithm. This algorithm transforms the original problem into an equality constrained problem by introducing two auxiliary vectors. In such a case, the CRO problem can be tackled by alternately solving several simple subproblems. Specifically, the subproblem w.r.t. the resource vector is convex, and the subproblems w.r.t. the auxiliary vectors are separable. Simulation results demonstrate that the proposed CRO scheme outperforms the traditional allocation schemes in terms of MTT performance. In addition, the performance of the CRO scheme is close to the optimal performance provided by the exhaustive method, but the computation load of the CRO scheme is lower than that of the exhaustive method. Finally, physical interpretations are presented to support our conclusions.

Computational microwave imaging with frequency-diverse metasurface (FDM) apertures is an emerging technology. In this article, we establish an experimental FDM microwave imaging system and address the waveform design problem based on the information theory, aiming to enhance the advantages of the FDM imaging. Two waveform design methods based on the different criteria are proposed for the FDM imaging system. The first waveform is designed by maximizing the mutual information between the object and the measured data with the constant transmitted energy, and the second one is designed by minimizing the transmitted energy while the mutual information is not less than a threshold. The performance of the proposed waveform design methods is evaluated by the data gathered by the self-established experimental FDM imaging system. The results show that the proposed waveform design methods are capable of improving the imaging quality or the imaging efficiency of the FDM imaging system.

For the problem of target localization under the multipath propagation environment, the existing methods are mainly restricted to the limited prior information of complex reflections, especially when the target is embedded in a mixed interference environment. They may suffer from performance degradation due to the shortage of target classification ability. To address this problem, we propose a target localization method based on iterative implementation with semiunitary constraint and eigen-decomposition technique, where a practical propagation scenario based on the spherical Earth model is considered. Compared to the previous works, the proposed method can automatically distinguish a real target from the mixed interference environment with improved localization accuracy. Neither additional decorrelation preprocessing nor prior information of the dynamic scenario is required. Both simulations and real data experiments validate the effectiveness and robustness of the proposed method.

Ultrawideband (UWB) radar can achieve ultrahigh-resolution inverse synthetic aperture radar (ISAR) imaging of noncooperative targets by transmitting UWB signals. However, the spatial-variant (SV) high-order migration through range cell (MTRC) and phase errors produced by the UWB radar system have seriously challenged the feasibility of conventional ISAR imaging algorithms. Moreover, maneuvering targets has exacerbated this problem compared with the steady ones. In this article, a UWB ISAR imaging algorithm of maneuvering targets with joint high-order motion compensation and azimuth scaling (JHOMCAS) is proposed. For the azimuth SV linear MTRC and 2-D SV high-order MTRC caused by the maneuvering rotational motion of the targets, the cascaded generalized keystone transform (GKT) is adopted for precise correction. It is worth noting that, when eliminating the SV MTRC by the cascaded GKT, the 2-D SV high-order phase errors induced by the maneuvering rotational motion must be accurately compensated, or MTRC correction will fail. The traditional autofocus methods usually only address the phase errors shared by the total target without due attention to the fine SV property. In response to this problem, this article first develops a joint 2-D SV autofocus and azimuth scaling algorithm (JSVAAS) to achieve the integration of SV high-order phase error compensation and azimuth scaling. A JHOMCAS algorithm is proposed to perform the joint processing of GKT and JSVAAS, “GKT-JSVAAS-GKT. ” This approach helps accomplish the high-precision UWB ISAR imaging of maneuvering targets, and the well-focused and scaled UWB ISAR images obtained will build a sound foundation for target classification and recognition. Extensive experiments based on both scattering point simulation data and electromagnetic calculation data verify that the proposed algorithm outperforms conventional ISAR imaging approaches in UWB ISAR imaging of maneuvering targets.

Under high sea state, ship targets make complex maneuvering motions due to strong disturbances such as sea waves and sea winds. Selecting the optimal imaging time interval to shorten the coherent processing interval (CPI) can reduce the complexity of motion errors, but the imaging resolution is affected as well, i.e., the motion error complexity and imaging resolution are constrained by each other. To address this problem, this article proposes a novel inverse synthetic aperture radar (ISAR) imaging framework for complex maneuvering ship targets to achieve the integration of super-resolution (SR) ISAR imaging and fine motion compensation (ISRFMC) under high sea state. With regard to the complex maneuvering motion of ship targets, we analyze the motion errors caused by the time-variant rotational velocity and imaging projection plane (IPP) on the echo signals, respectively, and a fine phase error model is established to uniformly represent the dual time-variant characteristic (DTVC) of complex maneuvering ship targets. Moreover, a deformed Akaike information criterion (DAIC) is developed to realize the adaptive selection of the phase error model with the image sharpness as the objective function. Underpinned by the Bayesian compressive sensing (BCS) theory, the SR ISAR imaging can be realized by solving a sparsity-driven optimization problem via a modified quasi-Newton solver. Particularly, the fine phase errors are constructed as the model errors of image reconstruction, and the particle swarm optimization algorithm (PSO) is utilized to solve the maximum image sharpness optimization problem in order to perform the joint fine motion compensation and azimuth scaling (JFMCAS). ISRFMC or the integration of SR ISAR imaging and fine motion compensation can be achieved through alternate iteration, so as to obtain well-focused and scaled high-resolution ISAR images of complex maneuvering ship targets under high sea state. Extensive experiments based on both simulated and real data verify that the proposed algorithm is capable of addressing the conflict between imaging resolution and motion error complexity under high sea state.

The high orbit height and long coherent processing interval (CPI) of geosynchronous (GEO) targets lead to the problems of ultralow signal-to-noise ratio (ULSNR) and complex signal modulation, posing great challenges to the traditional radar target detection and imaging algorithms. To address the problems, this article proposes a novel model-data codriven integration algorithm of detection and imaging for GEO targets with wideband radar. In this technique, underpinned by the transformation relationships between multiple spatial coordinate systems and the orbit prior information of GEO targets, we deduce the analytical expressions of the effective rotational vector of GEO targets so as to accomplish the model-driven optimal subaperture selection for integration of detection and imaging (OSASIDI). This considerably improves the processing performance and algorithm efficiency compared with traditional data-driven methods at ULSNR. In addition, we derive the radar equation of GEO targets for integration of detection and imaging in detail, which guides OSASIDI by analyzing the impacts of different parameters on detection and imaging performance. Aiming at the complex signal modulation problem caused by ultralong CPI (ULCPI) during the optimal subaperture (OSA) at ULSNR, we innovatively propose a model-data codriven integration of detection and imaging algorithm (MDCDIDI), which can eliminate the complex spatial-time-variant motion errors caused by the dual time-variant characteristic (DTVC) of effective rotational vector, so as to realize the focus-before-detection and obtain the well-focused inverse synthetic aperture radar (ISAR) images. Extensive experimental results from simulated data, which are generated from actual GEO parameters and the computer-aided-design (CAD) model of the Tiangong-I (TG-I) satellite, corroborate the effectiveness of the proposed algorithm.

Traditional radar waveforms are easily intercepted and interfered with by enemy’s reconnaissance system with the time–frequency periodic pattern recognition. Frequency and pulse repetition frequency (PRF) agility is an effective approach to decrease interception probability and increase anti-jamming capabilities. On the other hand, the agility brings about high sidelobes and the difficulty of parameter estimation using the range-Doppler signal processing. In this article, an optimization and high-resolution imaging algorithm for sparse stepped linear frequency modulation waveform (SSLFMW) with frequency and PRF agility is developed. The range and Doppler 2-D autocorrelation function of the agile waveform is investigated to pave a way to find an optimization strategy for frequency and PRF to suppress range and Doppler sidelobes. Relied on the pulse trains of low Doppler sidelobes, we propose a method of cognitive transmitting and motion retrieval based on the maximum likelihood principle to eliminate the frequency and range coupling in velocity estimation. The 2-D sparse reconstruction with conjugate gradient solver is proposed to efficiently reconstruct the high-resolution range-Doppler image with the frequency and PRF agility waveform. Both simulated and real-measured data sets are used to verify the improved performance of the proposal.

Due to hardware limitations, it is challenging for sensors to acquire images of high resolution in both spatial and spectral domains, which arouses a trend that utilizing a low-resolution hyperspectral image (LR-HSI) and a high-resolution multispectral image (HR-MSI) to fuse an HR-HSI in an unsupervised manner. Considering the fact that most existing methods are restricted by using linear spectral unmixing, we propose a nonlinear variational probabilistic generative model (NVPGM) for the unsupervised fusion task based on nonlinear unmixing. We model the joint full likelihood of the observed pixels in an LR-HSI and an HR-MSI, both of which are assumed to be generated from the corresponding latent representations, i.e., the abundance vectors. The sufficient statistics of the generative conditional distributions are nonlinear functions with respect to the latent variable, realized by neural networks, which results in a nonlinear spectral mixture model. For scalability and efficiency, we construct two recognition models to infer the latent representations, which are parameterized by neural networks as well. Simultaneously inferring the latent representations and optimizing the parameters are achieved using stochastic gradient variational inference, after which the target HR-HSI is retrieved via feedforward mapping. Though without supervised information about the HR-HSI, NVPGM still can be trained based on extra LR-HSI and HR-MSI data sets in advance unsupervisedly and processes the images at the test phase in real time. Three commonly used data sets are used to evaluate the effectiveness and efficiency of NVPGM, illustrating the outperformance of NVPGM in the unsupervised LR-HSI and HR-MSI fusion task.

Though widely studied, target detection in synthetic aperture radar (SAR) image is still a challenging problem. The classical convolutional neural network (CNN) samples spatial locations with the fixed geometric structure, and hence is incapable of learning the representations of varied-scale ships. It's difficult to locate multi-scale targets accurately in the complex scenes. On the other hand, to apply the classical models in SAR image, we need to duplicate the single-channel image to 3-channel one. The preprocess could not introduce the additional semantic information yet producing feature redundancy. To solve these problems, we introduced a new method for ship detection. We deployed the deformable convolutional block to learn features of ships with various scales at arbitrary locations. Different from the preceding works, the former shallow feature maps are also employed to enhance the representations of targets, especially small ships. In addition, the group normalization strategy is configured to alleviate internal covariate shift and accelerate the convergence. There is no need to make a tradeoff between the scale of model architecture and the batch size. Multiple comparative experiments on SSDD dataset are performed to demonstrate the advantage of proposed methods.

A colocated multiple-input multiple-output (MIMO) radar system may be interested in sending different data stream to multiple communication sinks widely separated in different directions with a physical layer security feature, i.e., a time-division dual function radar communications (DFRC) mode. Conventional channel precoding methods do not depend on the contents of the data streams. In this paper, given communication channels, we directly optimize constant-modulus transmit waveforms for a colocated MIMO radar, such that those communication sinks can decode the data stream in their own manner. Given the constellations of communication sinks, we design codes for different transmit antennas. Different from the precoding method, we transmit the optimized codes instead of precoded codes into space if a code combination needs to be transmitted into space. Meanwhile, the code optimization method enables to distribute different powers into space and can avoid eavesdroppers in other spatial directions from decoding the information transmitted. Numerical results with phase shift keying (PSK) constellations indicate that this method can form several intended code constellations in different spatial directions and the accuracy is higher if more transmit antennas are available. Meanwhile, the eavesdroppers in other space directions have a messed code constellations and it is hard to decode useful information from received signal.

Unmanned aerial vehicle (UAV) swarms have shown great potential for Internet of Things (IoT). Meantime, its malicious use may cause huge threat to the national security. UAV swarms show the characteristic of high density which poses formidable challenges to radar resolution in the defense of critical areas. In this article, we consider a radar equipped with the coprime array, and then, use the coherent long-time integration (LTI) technique and gridless sparse technique to detect and localize UAVs in a swarm. This strategy takes full account of advantages of the coprime array, coherent LTI technique, and gridless sparse technique, i.e.: 1) the coprime array can provide a larger array aperture than the uniform linear array with the same number of array elements to relieve the stress of the gridless sparse technique and 2) the combination of coherent LTI technique and gridless sparse technique can maximize their advantages and make up for their shortcomings. By mathematical analyses and extensive numerical examples, we show the superiority of the proposed strategy in terms of accurate detection and localization of UAV swarms.

The quality of inverse synthetic aperture radar (ISAR) images suffers seriously from the two-dimension (2-D) resolution and noise. The motion errors arising from translational and rotational motion further aggravate the image defocusing. For the limited bandwidth and short aperture (LB-SA) signal, this letter proposes a novel 2-D joint super-resolution (2D-JSR) ISAR imaging with joint motion compensation and azimuth scaling (JMCAS) algorithm. In this technique, a 2D-JSR signal model is established, enabling the 2-D high-resolution ISAR image to be generated by solving a sparsity-driven optimization problem with a modified quasi-Newton solver. In addition, a new JMCAS algorithm is developed to enhance the focusing performance of image. Not only can this algorithm jointly correct the range shift and phase error caused by translational motion, it can also complete the azimuth scaling and range spatial-variant phase error (RSVPE) compensation simultaneously. Through the iterative processing of 2D-JSR reconstruction and JMCAS, the well-focused and scaled high-resolution ISAR image can be obtained. Both simulated and real data experiments are provided to verify the effectiveness of the proposed algorithm.

To build a flexible and interpretable model for document analysis, we develop deep autoencoding topic model (DATM) that uses a hierarchy of gamma distributions to construct its multi-stochastic-layer generative network. In order to provide scalable posterior inference for the parameters of the generative network, we develop topic-layer-adaptive stochastic gradient Riemannian MCMC that jointly learns simplex-constrained global parameters across all layers and topics, with topic and layer specific learning rates. Given a posterior sample of the global parameters, in order to efficiently infer the local latent representations of a document under DATM across all stochastic layers, we propose a Weibull upward-downward variational encoder that deterministically propagates information upward via a deep neural network, followed by a Weibull distribution based stochastic downward generative model. To jointly model documents and their associated labels, we further propose supervised DATM that enhances the discriminative power of its latent representations. The efficacy and scalability of our models are demonstrated on both unsupervised and supervised learning tasks on big corpora.

The past years have witnessed a revival of neural network and learning strategies. These models configure multiple hidden layers hierarchically and require large amounts of labeled samples to estimate the model parameters. It is yet difficult to be met for target recognition under the realistic environments. For either space borne or airborne radars, collecting multiple samples with label information is very expensive and difficult. In addition, the huge computational cost and poor speed of convergence limit the practical applications. To address the problems, this article presents a new thought of receptive, under which a special hierarchy of feedforward neural network has been built. The proposed strategy consists of two sequential modules: 1) feature generation and 2) feature refinement. We first build pairwise baseline signals by means of the Riesz transform along the range and the azimuth, and extend them to a family of receptive signals using the bandpass filter bank. The input SAR image is then generally convoluted with the set of receptive signals to extract the global features. Certain kinds of information can be then exploited. We make the receptive signals predefined, rather than learned automatically, to handle the environment of a small sample size. In addition, the expert knowledge can be transmitted into the neural network. The resulting features are further refined by a special unit, wherein the input neurons and the latent states are bridged by the weights and the bias randomly generated. They are fixed during the training process. On the other hand, we cast the latent state into the Hilbert space, forming the kernel version of refinement. We aim to achieve the comparable or even better performance yet with limited training resources.

Thanks to the medium-to-high resolution and wide coverage imaging ability, satellite synthetic aperture radar (SAR) systems are momentously developed for intelligent maritime surveillance, especially for ship classification in SAR images. Previous researchers mostly utilized the geometric, radiometric, and structural features combined with the traditional machine-learning (ML) methods to conduct ship classification in high-resolution (HR) SAR images. However, the handcrafted features showed weak representation for the medium-resolution (MR) SAR images, and the normal ML methods are less powerful to deal with the intraclass diversity and interclass similarity in the MR SAR ship images. To address these issues, this article extends the dense convolutional networks to the MR SAR ship classification for better deep features extraction and proposes a multitask learning framework within which the softmax log-loss and the triplet loss are jointly minimized for more effective ship classification in MR SAR images. The triplet loss is computed by imposing a similarity constraint on the deep embeddings via the deep metric learning (DML) scheme such that the deep representations of the same class are pulled much closer to each other and those from different classes are pushed as farther apart as possible. However, the triplet loss in the usual DML utilizes the triplets mined in a training batch independently, which ignores the contextual information. So, we propose to impose a Fisher discrimination regularization term on the deep embeddings to explore the global information of learned embeddings, and hence the designed task-specific networks will be trained to have more robust and better recognition performance. Extensive experiments on the MR SAR ship data set collected from the Sentinel-1 SAR images demonstrate that the proposed method can achieve superior performance for the 3- and 5-class recognition tasks with regard to the comparing CNN models.

In the actual scene of interferometric inverse synthetic aperture radar (InISAR) imaging, the noncooperative targets may make a nonuniform 3-D rotational motion (3-D-RM), which contributes not only to the time-variant Doppler modulation but also to the spatial-variant wave path difference (SVWPD). This, in turn, seriously degrades the 3-D geometry reconstruction accuracy of the targets. Furthermore, it is an enormous challenge to realize InISAR imaging from sparse frequency band and sparse aperture (SFB-SA) signals. This article seeks to address the problems of fine image registration and 2-D joint sparse reconstruction (2-D-JSR) for InISAR imaging with SFB-SA signals. With regard to the maneuvering targets with 3-D-RM, a novel SVWPD signal model is established. Moreover, a new algorithm, named joint wave path difference compensation (JWPDC) algorithm, is developed to perform fine image registration. It can not only combine multiple channels to achieve image registration but also jointly compensate for the non-SVWPD (NSVWPD) and SVWPD. A joint multichannel 2-D-JSR (JMC-2-D-JSR) ISAR imaging algorithm is also proposed according to the SFB-SA signal model to produce high-resolution ISAR images. Underpinned by the Bayesian compressive sensing (BCS) theory, the JMC-2-D-JSR ISAR imaging can be realized by solving a sparsity-driven optimization problem via a modified quasi-Newton solver. Through iterative processing of JMC-2-D-JSR and JWPDC, the high-quality 3-D InISAR images of maneuvering targets with 3-D-RM can be obtained. Extensive experimental results based on both simulated and real data corroborate the effectiveness of the proposed algorithm that outperforms other available InISAR imaging frameworks in 2-D imaging, 3-D imaging, and motion compensation.

Various kinds of data are generated from industrial Internet of Things, and these data can be applied for connecting production equipment, identifying and locating items, etc. These data should be forwarded to the decision center for further analyses, especially in wartime. Thus, the channel status information (CSI) for industrial big data transmission has to be acquired. In this article, we develop a system architecture for industrial big data (BD) transmission based on radar-communication integration with arbitrary geometrical array. The traditional channel estimation method, which usually utilizes the regular antenna array to estimate the CSI, cannot be applied to the arbitrary geometrical array. Here, we use the manifold separation technique to transform the complex array configuration into regular array and the downlink channel covariance matrix is estimated by exploiting the frequency calibration technique when the uplink channel covariance matrix is received. The computational complexity for the proposed method and other state-of-the-art methods are analyzed. The simulation results prove that the proposed method can achieve excellent estimation performance for its application in radar-communication integration.

Unmanned aerial vehicle (UAV) swarms-enabled mobile edge computing system can be deployed in critical industrial zones for monitoring. Meanwhile, its malicious use may bring great threat to the security, and the accurate detection, and localization are important. UAV swarms show characteristics of the high density, small radar cross section, far range, and time-varying motion, and have posed formidable challenges to the accurate detection and localization. In this article, the accurate detection and localization of UAV swarms are investigated, and an effective method is proposed based on the Dechirp-keystone transform, and frequency-selective reweighted trace minimization. It inherits high robustness of the coherent long-time integration technique and superresolution of the gridless sparse technique. Mathematical analyzes and numerical simulations validate its superiorities in accurate detection and localization of UAV swarms.

We propose a meta network (MNet) for the problem of target-aspect missing in radar high-resolution range profile (HRRP)-based noncooperative target recognition, where a classifier must be generalized to new aspects not seen in the training set, given only a small number of HRRP data of each new aspect. The MNet is a time domain convolutional neural network (TCNN) that is built based upon recent progress in meta-learning. In effect, it learns a model that is easy and fast to fine-tune, allowing the adaptation to happen in the right space for fast learning. Besides, we construct a new controllable HRRP dataset suitable for the scenario of noncooperative target-aspect missing using electromagnetic simulation. Compared with the traditional methods, the MNet is more efficient and could achieve better performance. Extensive experiments on the simulated HRRP dataset are conducted to illustrate the effectiveness of the proposed method.

Synthetic aperture radar (SAR) images are especially susceptible to the target aspect angles. For the SAR target recognition, the lack of training data at different aspect angles inevitably deteriorates the performance. To solve the problem, this letter introduces an angular rotation generative network (ARGN). It is actually an attribute-guided transfer learning method, and the shared attribute between the source and target domains is the target aspect angle. The aspect angle of the data for each type in the source domain covers in the range of 0°-360°, while the information of the aspect angle in the target domain is not complete. Assume that there is a mapping in the feature space between the two images of the same target under different aspect angles, and the mapping relation is learned from sufficient data in the source domain. Then, the mapping can also be applied in the target domain according to the idea of transfer learning. The learned knowledge contained in the feature space helps to improve the target recognition performance in the target domain. Experimental results on the moving and stationary target acquisition and recognition (MSTAR) benchmark data set illustrate that our framework is efficient. For the three-class recognition of the MSTAR data set, the recognition rate is about 87% even when only 23 samples of each class are utilized as the training data. For a given target data, the generation results with counterclockwise rotation of 1°-90° for the aspect angle are performed. The qualitative and quantitative comparisons between the generated images and the real data are also displayed.

In a multipath propagation environment, prevalent target localization methods are mainly based on the classical two-ray propagation model without considering other reflected waves. Because the received target echoes are considerably corrupted by multipath reflections in the case of complex terrain, these prevalent methods might fail to work or achieve poor performance. To solve this problem, we first consider a practical multipath propagation scenario to reveal the dynamic structural relationship of the spatial paths based on the spherical earth model. Subsequently, a target localization algorithm based on low-rank decomposition (LRD) and sparse representation (SR) framework is proposed. The proposed algorithm can effectively mitigate the effects of complex multipath interference without using any prior knowledge on the illuminated terrain and the reflecting paths. Experiments on synthetic data and real data validate the effectiveness of the proposed algorithm.

With regard to the multifunction radar transmitting sparse stepped-frequency-modulation (SSFM) signal for inverse synthetic aperture radar (ISAR) imaging, the received echo signal is usually sparse in two dimensions, i.e., sparse stepped-frequency-modulation and sparse aperture waveforms (SSFM-SAWs), and there are translational and rotational motion errors between subpulses. The two problems seriously challenge the feasibility of conventional 1-D sparse reconstruction algorithms. This article proposes a novel high-resolution ISAR imaging and motion compensation with the 2-D joint sparse reconstruction (2D-JSR) algorithm. In this technique, a 2D-JSR dictionary is established according to the SSFM-SAW signal model. Based on the Bayesian compressive sensing (BCS) theory, the 2D-JSR is then transformed into solving a sparsity-driven optimization problem with l1-norm constraint. With the accommodation of a modified quasi-Newton solver, the exact recovery of SSFM-SAW can be achieved. In addition, a new algorithm, named joint translational motion compensation and range spatial-variant autofocus (JTSVA) algorithm, is also developed to realize motion parameters by a two-step estimation. Integrating with 2-D coupling information of echo signal and the efficient and robust motion compensation algorithm, the accurate motion parameters together with well-focused and scaled high-resolution ISAR images can be obtained. Extensive experiments based on both simulated and real data demonstrate that the proposed algorithm is capable of the precise reconstruction of ISAR images and the effective suppression of both motion errors and noise.

Accurate ranging and wideband tracking are treated as two independent and separate processes in traditional radar systems. As a result, limited by low data rate due to nonsequential processing, accurate ranging usually performs low efficiency in practical application. Similarly, without applying accurate ranging, the data after thresholding and clustering are used in wideband tracking, leading to a significant decrease in tracking accuracy. In this article, an integrated Kalman filter of accurate ranging and tracking is proposed using methods of phase-derived-ranging and Bayesian inference in wideband radar. Besides the motion state, in this integrated Kalman filter, the complex-valued high-resolution range profile (HRRP) is also introduced as a reference signal by coherent integration in a sliding window, which incorporates target's scattering distribution and phase characteristics. Corresponding kinetic equations are derived to predict the motion state and the reference signal in the next moment. A ranging process is constructed based on the received signal and the predicted reference signal in order to estimate innovation using methods of phase-derived-ranging and Bayesian inference, and a sequential update for motion state can be accomplished with the Kalman filter as well. In every recursion, the complex-valued reference signal is also updated by coherently integrating the latest pulses. The integrated Kalman filter takes full use of high range resolution and phase information, improving both efficiency and precision compared with conventional approaches of ranging and wideband tracking. Implemented in a sequential manner, the integrated Kalman filter can be applied in a real-time application, realizing simultaneous ranging with high precision and wideband tracking. Finally, simulated and real-measured experiments confirm the remarkable performance.

The stepped-frequency waveform is usually used to synthesize a wideband signal in the radar imaging system. To reduce the amount of data and coherent pulse intervals (CPIs), as well as to improve antijamming abilities, the sparse stepped frequency is employed in the area of inverse synthetic aperture radar (ISAR) imaging. Nevertheless, the traditional sparse stepped linear frequency modulation waveform (SSLFMW) has a shortage of high grating lobes caused by missing frequency bands, resulting in a degradation of the ISAR imaging quality. Many methods have been proposed to reduce the effect of grating lobes by echo signal processing. However, the method of grating lobe reduction is rarely studied from the aspect of waveform optimization. In this article, a novel SSLFMW with the low grating lobes in the ISAR imaging system is proposed. By deriving the autocorrelation function (ACF), the relation between grating lobes and waveform parameters, including stepped-frequency and phase-coded elements, is established. An optimization method based on alternate iteration is designed to optimize waveform parameters and reduce grating lobes. Based on this optimized SSLFMW, we establish an ISAR imaging framework with the compressive sensing (CS) theory. Finally, the experiments are designed to show that the optimized SSLFMW has lower grating lobes. Both simulated and real measured data are used to prove that the optimized waveform has a better performance in the high-resolution range profile (HRRP) synthesis and ISAR imaging compared with the traditional SSLFMW.

Due to hardware limitations of the imaging sensors, it is challenging to acquire images of high resolution in both spatial and spectral domains. Fusing a low-resolution hyperspectral image (LR-HSI) and a high-resolution multispectral image (HR-MSI) to obtain an HR-HSI in an unsupervised manner has drawn considerable attention. Though effective, most existing fusion methods are limited due to the use of linear parametric modeling for the spectral mixture process, and even the deep learning-based methods only focus on deterministic fully-connected networks without exploiting the spatial correlation and local spectral structures of the images. In this paper, we propose a novel variational probabilistic autoencoder framework implemented by convolutional neural networks, in order to fuse the spatial and spectral information contained in the LR-HSI and HR-MSI, called FusionNet. The FusionNet consists of a spectral generative network, a spatial-dependent prior network, and a spatial-spectral variational inference network, which are jointly optimized in an unsupervised manner, leading to an end-to-end fusion system. Further, for fast adaptation to different observation scenes, we give a meta-learning explanation to the fusion problem, and combine the FusionNet with meta-learning in a synergistic manner. Effectiveness and efficiency of the proposed method are evaluated based on several publicly available datasets, demonstrating that the proposed FusionNet outperforms the state-of-the-art fusion methods.

To solve target recognition under extended environments, this paper proposes sparse representation in the transformed domain. Since the signal energy in the frequency domain is mainly concentrated on a small portion of low frequencies, this part of spectrum therefore carry the vital information that distinguishes a class of target from the other. We intend to define a frequency descriptor by the bag of low frequencies. The defined descriptor is used to build sparse signal modeling. The frequency descriptors of the training are concatenated to form an over-complete dictionary. It is used to encode the counterpart of query as a linear combination of themselves. Sparsity has been harnessed to generate the optimal representation, from which the inference can be reached.

For desired resolution, high-squint SAR has a large coherent processing interval (CPI). In this case, the high signal-to-clutter ratio (SCR) echo of moving target can not be guaranteed due to the presence of ground clutter, and the maneuvering motion of moving target usually causes high-order phase terms, which can not be neglected for precise focusing. In this paper, the climb method is applied to the range-compressed echo containing moving target to improve the SCR. For the subsequent focusing, we assume that the target of interest has constant velocity in subaperture CPI, but maneuvering motion parameters for the whole CPI. Within the short subaperture CPI, the target signal can be simplified as a three-order polynomial function with the unknown coefficients to be estimated by some time-frequency analysis tools, including Hough transform and fractional Fourier transform. Then the subaperture Doppler parameters are combined to form a total least square problem, outputting the high-order phase terms. The effectiveness of the proposed GMTIm method is validated by real-measured high-squint SAR data.

Artificially structured metamaterials apertures antennas (MAA) enables producing the pseudorandom and spatially variant radiation fields to encode spatial information and retrieve scene images using computational imaging (CI) algorithms. Combined with synthetic aperture radar (SAR) technologies, by moving a linear shape MAA in crosswise direction, a hybrid imaging system with the combination of MAA and SAR is demonstrated in this paper. Focusing on this peculiar imaging geometry, we propose a postprocessing algorithm which combines the classic omega-k algorithm and CI algorithms to achieve fully three dimensional scene images. Compared with traditional frequency-diverse imaging reconstruction algorithms, the postprocessing is more efficient could achieve as high efficiency as SAR algorithms do. Extensive imaging simulations are conducted to illustrate the effectiveness of the proposed algorithms.

In this paper, a unified Bayesian max-margin discriminant projection framework is proposed, which is able to jointly learn the discriminant feature space and the max-margin classifier with different relationships between the latent representations and observations. We assume that the latent representation follows a normal distribution whose sufficient statistics are functions of the observations. The function can be flexibly realized through either shallow or deep structures. The shallow structure includes linear, nonlinear kernel-based functions, and even the convolutional projection, which can be further trained layerwisely to build a multilayered convolutional feature learning model. To take the advantage of the deep neural networks, especially their highly expressive ability and efficient parameter learning, we integrate Bayesian modeling and the popular neural networks, for example, mltilayer perceptron and convolutional neural network, to build an end-to-end Bayesian deep discriminant projection under the proposed framework, which degenerated into the existing shallow linear or convolutional projection with the single-layer structure. Moreover, efficient scalable inferences for the realizations with different functions are derived to handle large-scale data via a stochastic gradient Markov chain Monte Carlo. Finally, we demonstrate the effectiveness and efficiency of the proposed models by the experiments on real-world data, including four image benchmarks (MNIST, CIFAR-10, STL-10, and SVHN) and one measured radar high-resolution range profile dataset, with the detailed analysis about the parameters and computational complexity.

Enhancing the contrast of target and clutter is a crucial issue in synthetic aperture radar (SAR) image target detection. In this paper, we define a novel subspace, called minimal clutter-to-signal ratio (MCSR) subspace, which can minimize the clutter-to-signal ratio (CSR) by projecting the feature vector to the subspace. Based on MCSR, a novel polarimetric contrast enhancement method is proposed. The MCSR subspace is learned based on the commonly used polarimetric feature vectors extracted from the labeled training SAR image pixels. The feature vectors extracted form candidate SAR image pixels are projected to the MCSR subspace. By calculating the square norm of each transformed feature vector, an enhanced image can be obtained. It is demonstrated that the existing optimization of polarimetric contrast enhancement (OPCE) is a special case of the proposed method to some extent. Experimental results show that our method outperforms the traditional OPCE method on the RadarSat-2 SAR data.

For modern meter-wave radar, the performance of low-angle target altitude measurement is limited by multipath phenomenon, especially in the complex terrain environment where the multipath signal is perturbed by irregular surface. To address this problem, a practical signal model for meter-wave radar in practical terrain is first presented, where the influence of the perturbed multipath caused by irregular reflecting surface is taken into consideration. A novel compressive sensing (CS) based altitude measurement algorithm, combined with alternative optimization and dictionary updating techniques, is then proposed, in which the perturbation caused by the complex terrain can be iteratively compensated to estimate the target altitude more precisely. Numerical results based on both simulated data and real data demonstrate the effectiveness of the proposed algorithm under complex terrain environment.

To detect ships robustly and automatically in monitoring the marine areas, polarimetric synthetic aperture radar imagery is more and more important. In this letter, three superpixel-level dissimilarity measures are developed to enhance the contrast between ship targets and sea clutter, which are then used to construct an automatic detection algorithm. In the proposed method, multiscale superpixels are first generated. Second, the measurements between a certain superpixel and surrounding ones are calculated. The dissimilarity measures are then transformed from the superpixel level to the pixel level. Third, kernel Fisher discriminant analysis is utilized to improve the separability between ship targets and clutter. Finally, linear support vector machine classifier is utilized to complete the detection automatically. Experiments on the synthetic and real data demonstrate that the proposed method is effective for ship detection with only few false alarms existing, especially under the low signal-to-clutter ratio.

An important step in multi-sensor data fusion is sensor registration, namely, to estimate sensors' range and azimuth biases from their asynchronous measurements. Assuming the target moves in a straight line with an unknown constant velocity, we propose a two-stage nonlinear least square (LS) approach to this problem. More specifically, in stage I, each sensor first estimates its own range bias individually, and then in stage II, all sensors jointly estimate their azimuth biases. We show that both of the nonconvex LS problems can be solved to global optimality under mild conditions. Simulation results show that the root mean square error (RMSE) of the proposed approach is quite close to the Cramér-Rao lower bound (CRLB) when the level of the measurement noise is small.

It is challenging to develop stochastic gradient based scalable inference for deep discrete latent variable models (LVMs), due to the difficulties in not only computing the gradients, but also adapting the step sizes to different latent factors and hidden layers. For the Poisson gamma belief network (PGBN), a recently proposed deep discrete LVM, we derive an alternative representation that is referred to as deep latent Dirichlet allocation (DLDA). Exploiting data augmentation and marginalization techniques, we derive a block-diagonal Fisher information matrix and its inverse for the simplex-constrained global model parameters of DLDA. Exploiting that Fisher information matrix with stochastic gradient MCMC, we present topic-layer-adaptive stochastic gradient Riemannian (TLASGR) MCMC that jointly learns simplex-constrained global parameters across all layers and topics, with topic and layer specific learning rates. State-of-the-art results are demonstrated on big data sets.

In this letter, the local scattering mechanism difference based on regression kernel (LSMDRK) is developed as a discriminative feature for ship detection. The LSMDRK measures the scattering mechanism dissimilarity of a center pixel to its neighboring pixels. A ship detection scheme is proposed based on the LSMDRK. The detection scheme consists of two stages. In the feature extraction stage, polarimetric target decomposition is required to improve the discriminative ability of the descriptor. In the detection stage, a saliency detection strategy is utilized to construct the saliency map. Then, local maximum detection is employed. Finally, an adaptive threshold method is designed to achieve the final detection. The effectiveness of the detection scheme is validated by a RADARSAT-2 data set. Experimental results demonstrate that the proposed method can acquire a better detection on weak targets and has much less false alarms than some classical detection methods.

To improve the synthetic aperture radar (SAR) target discrimination performance under complex scenes, this letter presents a new SAR target discrimination method based on the bag-of-words model. The method contains three main stages. In the local feature extraction stage, the SAR-SIFT feature is extracted. In the feature coding stage, we improve the existing category-specific and shared dictionary learning (CSDL) and propose the sample-reweighted CSDL (SR-CSDL). The local features are sparsely coded using the codebook learned from SR-CSDL. In the feature pooling stage, spatial pyramid matching with max pooling is used to aggregate the local coding coefficients to generate the global feature for each chip image. Experimental results using the miniSAR data verify the effectiveness of the proposed method.

Target discrimination has been one of the hottest issues in the interpretation of synthetic aperture radar (SAR) images. However, the presence of speckle noise and the absence of robust features make SAR discrimination difficult to deal with. Recently, convolutional neural network has obtained state-of-the-art results in pattern recognition. In this letter, we propose a target discrimination framework that jointly uses intensity and edge information of SAR images. This framework contains three parts, namely, feature extraction block, feature fusion block, and final classification block. In addition, a novel feature fusion method that can preserve the spatial relationship of different features is introduced. Experimental results on the miniSAR data demonstrate the effectiveness of our method.

In this paper, we propose a new discriminative non-linear dictionary learning approach, called correlation constrained structured kernel KSVD, for object recognition. The objective function for dictionary learning contains a reconstructive term and a discriminative term. In the reconstructive term, signals are implicitly non-linearly mapped into a space, where a structured kernel dictionary, each sub-dictionary of which lies in the span of the mapped signals from the corresponding class, is established. In the discriminative term, by analyzing the classification mechanism, the correlation constraint is proposed in kernel form, constraining the correlations between different discriminative codes, and restricting the coefficient vectors to be transformed into a feature space, where the features are highly correlated inner-class and nearly independent between-classes. The objective function is optimized by the proposed structured kernel KSVD. During the classification stage, the specific form of the discriminative feature is needless to be known, while the inner product of the discriminative feature with kernel matrix embedded is available, and is suitable for a linear SVM classifier. Experimental results demonstrate that the proposed approach outperforms many state-of-the-art dictionary learning approaches for face, scene, and synthetic aperture radar vehicle target recognition.

In this paper, a novel noise-robust recognition method for high-resolution range profile (HRRP) data is proposed based on target scatterer pattern to enhance its recognition performance under the test condition of low SNR. The target dominant scatterers are first extracted based on the scattering center model of complex HRRP data via the orthogonal matching pursuit (OMP) algorithm to realize noise reduction. Then a scatterer matching recognition algorithm based on Hausdorff distance (HD) is developed with the magnitudes and locations of extracted dominant scatterers used as the feature patterns. Experimental results on the measured HRRP data demonstrate that the proposed method can improve the recognition performance under the relatively low SNR condition for both orthogonal and superresolution representations of scattering center model.

The problem of detecting a subspace signal is studied in colored Gaussian noise with an unknown covariance matrix. In the subspace model, the target signal belongs to a known subspace, but with unknown coordinates. We propose a modified Rao test (MRT) by introducing a tunable parameter. The MRT is more general, which includes the Rao test and the generalized likelihood ratio test as special cases. Moreover, closed-form expressions for the probabilities of false alarm and detection of the MRT are derived. Numerical results demonstrate that the MRT can offer the flexibility of being adjustable in the mismatched case where the target signal deviates from the presumed signal subspace. In particular, the MRT provides better mismatch rejection capacities as the tunable parameter increases.

Multiple scattering effects are commonly ignored in the detection and estimation of scatterers in signal processing research, because the energy of the first-order scattering is much larger than that of higher-order components. Although multiple scattering can significantly increase the estimation precision of point scatterers, it does not always lead to an improvement. Identifying conditions under which multiple scattering is beneficial or detrimental to estimation in a general setup is still an open problem. In this paper, we consider the effects of multiple scattering on the localization of two point scatterers. By comparing the Fisher information matrix on location parameters when multiple scattering exists and does not exist, we show analytically that information on ranges can benefit estimating directions of arrival via multiple scattering when the two scatterers are in far-field and well resolved.

Microwave metamaterial aperture imaging radar (MMAIR) is capable of generating high resolution images without using mechanical scanning or antenna arrays. In MMAIR, metamaterial elements are specifically embeded into a parallel plate waveguide, whose resonance frequencies vary among a wide bandwidth. Different radiation fields are gained by wide-band waveforms, and the scene information are measured by measurement matrix which consists of a set of radiation modes. According to the Compressed Sensing theory, MMAIR performance is restricted by the limited frequency measurement modes. Herein we propose a rotation-synthesis approach to synthesize radiation measurement modes. By rotating the metamaterial aperture panel around the panel axis, the approach exploits the radiation field's multi-beam spatial diversity, and with the azimuth rotation, the radiation field pattern varies relative to the scene under the radiation field. As a result, significant MMAIR imaging enhancement is achieved with the crucial increase of radiation measurement modes. The simulations verify the effectiveness and image improvement of the proposed method.

Many methods have been proposed to improve the performance of synthetic aperture radar (SAR) target recognition but seldom consider the issues in real-world recognition systems, such as the invariance under target translation, the invariance under speckle variation in different observations, and the tolerance of pose missing in training data. In this letter, we investigate the capability of a deep convolutional neural network (CNN) combined with three types of data augmentation operations in SAR target recognition. Experimental results demonstrate the effectiveness and efficiency of the proposed method. The best performance is obtained by using the CNN trained by all types of augmentation operations, showing that it is a practical approach for target recognition in challenging conditions of target translation, random speckle noise, and missing pose.

In this letter, a new superpixel-based constant-false-alarm-rate (CFAR) target detection algorithm for high-resolution synthetic aperture radar (SAR) images is proposed. The detection algorithm consists of three stages, i.e., segmentation, detection, and clustering. In the segmentation stage, a superpixel-generating algorithm is utilized to segment the SAR image. In the detection stage, based on the superpixels generated, the clutter distribution parameters for each pixel can be adaptively estimated, even in the multitarget situations. Then, the two-parameter CFAR test statistic can be adopted for detection. In the clustering stage, the hierarchical clustering is used to cluster the detected superpixels to get the candidate targets. The effectiveness of the proposed algorithm is demonstrated using the miniSAR data.

A Bayesian classifier for sparsity-promoting feature selection is developed in this paper, where a set of nonlinear mappings for the original data is performed as a pre-processing step. The linear classification model with such mappings from the original input space to a nonlinear transformation space can not only construct the nonlinear classification boundary, but also realize the feature selection for the original data. A zero-mean Gaussian prior with Gamma precision and a finite approximation of Beta process prior are used to promote sparsity in the utilization of features and nonlinear mappings in our model, respectively. We derive the Variational Bayesian (VB) inference algorithm for the proposed linear classifier. Experimental results based on the synthetic data set, measured radar data set, high-dimensional gene expression data set, and several benchmark data sets demonstrate the aggressive and robust feature selection capability and comparable classification accuracy of our method comparing with some other existing classifiers.

To improve the target detection performance under a low signal-to-clutter ratio, this letter presents a new polarimetric synthetic aperture radar (PolSAR) ship detector based on superpixel-level scattering mechanism (SM) distribution features. The proposed method is based on the observation that the SMs of targets and clutter have different distributions in the classical H/α plane. To make use of this difference in ship detection, multiscale superpixels are first generated for PolSAR images. Then, two features describing the SM distribution in the superpixel are proposed. Based on these features, a test statistic independent of the scattering intensity is finally defined. The performance improvement of the proposed method is verified using a synthetic data set and real PolSAR images obtained from a RADARSAT-2 data set.

Our aim in this work is to analyze the detection performance of a constant false alarm rata matched filter (CFAR-MF) which was developed for the detection problem in white Gaussian noise with unknown noise power. An exact expression for the detection probability of the CFAR-MF is derived in the mismatched case where mismatch exists between the actual signal steering vector and the nominal one. This theoretical expression can be used to facilitate the performance evaluation of the CFAR-MF in real-world scenarios when signal mismatch cannot be neglected.

In this paper, we introduce a new max-margin discriminant projection method, which takes advantage of the latent variable representation for support vector machine (SVM) as the classification criterion. Specifically, the proposed model jointly learns the discriminative subspace and classifier in a Bayesian framework by conditioning on augmented variables. Moreover, an extended nonlinear model is developed based on the kernel trick, where the similar model can be used in this setting with few modifications. To explore the sparsity in the kernel expansion, we use the spike-and-slab prior to seek basis vectors (BVs) from the corresponding candidates. Unlike existing methods, which employ BVs to approximate the original feature space, in our method BVs are sought to associate the final classification task. Thanks to the conditionally conjugate property, the parameters in our models can be inferred via the simple and efficient Gibbs sampler. Finally, we test our methods on synthesized and real-world data, including large-scale data sets to demonstrate their efficiency and effectiveness.

This letter presents an adaptive inverse synthetic aperture radar (ISAR) imaging algorithm based on evidence framework. This imaging algorithm iterates between sparse coding and parameter estimation until a fixed number of iterations is reached. An efficient Hessian update scheme is applied for sparse coding. In the parameter estimation step, the closed forms of parameters are obtained based on evidence framework. Experiments based on simulated and measured data are given to show the efficiency of the proposed ISAR imaging algorithm.

A novel inverse synthetic aperture radar (ISAR) imaging algorithm for micromotion targets based on multiple sparse Bayesian learning (MSBL) is proposed. First, the signal of the main body is reconstructed by the MSBL method based on the property of its common profile. Subsequently, the signal of rotating parts can be obtained by removing the main body signal from the original signal. Finally, a clear ISAR image of the main body and the micromotion parameter of the rotating parts can be obtained. Numerical results based on simulated and measured data show that the proposed algorithm can not only acquire a clear ISAR image of the main body but also extract the micromotion parameter of the rotating parts effectively.

The triplet Markov field (TMF) can obtain more promising classification results of nonstationary images than the Markov random field (MRF). However, TMF has limitedly specialized applications to polarimetric synthetic aperture radar (PolSAR) images with nonstationarity properties. In addition, it is difficult to interpret the meaning of the auxiliary field derived by TMF. This implies that the auxiliary field may not have the physical meaning. We propose Wishart TMF with a specific auxiliary field for PolSAR image classification. We define a smoothness characteristic, which describes the extent of pixel smoothness in its neighborhood. This characteristic acts on the energy of the proposed TMF to supervise the classification of the auxiliary field. The auxiliary field can distinguish the smoothness stationarity and nonsmoothness stationarity of PolSAR images, which indicates that the auxiliary field has the specific physical meaning. The effectiveness of the proposed TMF is demonstrated by real PolSAR image classification experiments.

The motivation of this letter is to resolve the nonnegative eigenvalue constraint (NNEC) problem of four-component decomposition (FCD). It is analyzed that the NNEC is an essential requirement for remainder coherency matrices in the FCD, however the measured polarimetric synthetic aperture radar (POLSAR) data experiment shows there exits the NNEC problem that some remainder coherency matrices of the FCD do not satisfy the NNEC, which means these matrices are not positive semi-definite. In addition, it is analyzed that the scheme using the nonnegative eigenvalue decomposition (NNED) for three-component decomposition (TCD) cannot be directly extended to the FCD to overcome the NNEC problem, so a scheme using the NNED for the FCD is proposed as follow. From matrix theory, we draw a conclusion that if the last remainder coherency matrix satisfies the NNEC, then all remainder coherency matrices also satisfy the NNEC; we successively analyze that the NNEC problem of the last remainder coherency matrices results from the overestimation of scattering powers. Then a shrinkage coefficient is used to depress all possible overestimations of scattering powers, and the overestimation case with the minimum remainder power is chosen to resolve the NNEC problem. Moreover, we have simplified the solution to NNED, which is used to calculate the shrinkage coefficient. The measured POLSAR data experiment shows that the proposed FCD can further enhance double-bounce scattering and depress volume scattering for urban areas.

Recently, compressive sensing (CS) has been successfully used in inverse synthetic aperture radar (ISAR) imaging. Since the exact sparse reconstruction, i.e., l0-norm constraint, is NP hard, l1-norm relaxation is widely used at the cost of performance degradation in the sparseness of the solution. The performance of existing CS-based ISAR imaging algorithms is sensitive to the regularized factor, which should be adjusted manually. This makes the existing algorithms inconvenient to be used in practice. It is well known that sparse Bayesian learning (SBL) acts as an effective tool in regression and classification, which is closely related to the CS. Furthermore, all the necessary parameters can be estimated using an efficient evidence maximization procedure in SBL, which retains a preferable property of the l0-norm diversity measure and can give more sparse solution. Motivated by that, a fully automated ISAR imaging algorithm based on SBL is proposed in this paper. Experimental results based on simulated and measured data show that the proposed algorithm keeps a better balance between the computation load and the sparsity of the reconstruction signal than the existing algorithms.

This paper presents a new model-based decomposition method for the polarimetric synthetic aperture radar coherency matrices. We improve the model flexibility from the following two aspects: To reach a compromise between model flexibility and computation complexity, for the volume scattering component, the elementary scatterer shape is allowed to change from sphere/flat plate to dipole, then to dihedral, whereas orientation randomness is simplified by only considering two cases. Different orientation angles are considered for each component. Since the models become more complex, new decomposition procedures are developed. The three-component decomposition is first reformulated as a constrained sparse representation problem. Then, inspired by the orthogonal matching pursuit variant developed by Bruckstein et al. in 2008, new decomposition procedures are designed. The effectiveness of the proposed method is verified using a synthetic data set and two real SAR data sets, including a RADARSAT-2 data set and the NASA/JPL AIRSAR data set over San Francisco Bay.

In this letter, we introduce a novel noise-robust modification method for Gaussian-based models to enhance the performance of radar high-resolution range profile (HRRP) recognition under the test condition of low signal-to-noise ratio (SNR), and we develop an efficient scheme for its computation. This noise-robust modification method is implemented by revising the trained Gaussian-based model according to the estimated SNR of test HRRP. We apply the proposed method to adaptive Gaussian classifier and truncated stick-breaking hidden Markov model. Experimental results demonstrate that the proposed method can significantly improve the average recognition rate for noisy HRRP test samples while offering recognition performance comparable to that of original model for clean HRRP test samples. Moreover, even when the SNR of test HRRP samples is not precisely estimated, we can still obtain an acceptable result with the proposed method.

A novel method is proposed for moving wheeled vehicle and tracked vehicle classification using micro-Doppler features from returned radar signals within short dwell time. In this method, an adaptive analysis technique called Empirical Mode Decomposition (EMD) is utilized to decompose the motion components of moving vehicles, and a hierarchical classification structure using the decomposition results of returned signals is proposed to discriminate the two kinds of vehicles. The first stage of the structure elementarily identifies the tracked vehicle data by checking the existence of its unique feature and a further classification via our proposed features based on EMD is implemented in the second stage by using Support Vector Machine (SVM) classifier. Experimental results based on the simulated data and measured data are presented, including the performance analysis for low signal-to-noise ratio (SNR) case, generalization evaluation for different target circumstances and comparison with some related methods.

This paper presents a new hierarchical scheme for detecting ships from high-resolution synthetic aperture radar (SAR) images. The scheme consists of two stages: detection and discrimination. In the detection stage, the existing internal Hermitian product is extended to obtain a new detector. The new detector makes a combined use of the complex coherence among more than two subapertures and the intensity of each subaperture. When the subaperture number is increased, the target/clutter contrast is shown to be improved. Ship candidates are obtained by applying a threshold. Ship discrimination is performed by using one-class classification. The covariance descriptor, developed by Tuzel in 2006, is introduced to SAR ship discrimination as the feature. The traditional one-class quadratic discriminator is used as the discriminator. After this stage, most false alarms are rejected, and the real ship targets in the candidates are maintained. The effectiveness of the proposed scheme is verified using RADARSAT-2 data. Experimental results show that the proposed scheme can detect most ship targets in the image and few false alarms occur.

Radar high-resolution range profiles (HRRPs) are typical high-dimensional non-Gaussian and inter-dimensional dependently distributed data, the statistical modelling of which is a challenging task for HRRP based target recognition. Considering the inter-dimensional dependence, a recent work applied Factor Analysis (FA) to model radar HRRP data and showed promising recognition results, which however still restricts to Gaussian distribution. This paper aims to simultaneously consider the inter-dimensional dependence and the non-Gaussian distribution, by using Local Factor Analysis (LFA) model. For not only learning parameters but also appropriately selecting the component number and local hidden dimensionalities, we adopt the automatic Bayesian Ying-Yang (BYY) harmony learning, in order to relieve the extensive computation and inaccurate evaluation encountered in the conventional two-phase implementation. Moreover, a heuristic aspect-frame partition is implemented based on the BYY harmony criterion rather than AIC or BIC in the previous work, to tackle the radar HRRP's target-aspect sensitivity. Experiments show improved recognition performances over on the same measured HRRP dataset, i.e., for both equal interval and heuristic aspect-frame partitions, LFA automatically learned by BYY always outperforms FA selected by a two-phase procedure with either AIC or BIC.

In this paper, we propose an iterative design approach to jointly optimize probing signal waveforms and a receive filter bank for a multiple-input multiple-output (MIMO) radar under a constant modulus constraint. The design goals are to approximate a desired beampattern and to minimize the auto-/cross- correlation levels of the probing signal waveforms for different time lags and between different spatial angles. Since the overall design problem is nonconvex, we propose to optimize the transmit probing signals and receive filter bank separately and alternately. The optimization of receive filter bank is a standard least squares problem, while the optimization of the constant modulus transmit signal waveforms is a norm-constrained least squares problem which can be approximately solved using a low-rank semidefinite relaxation procedure. We demonstrate the effectiveness of our proposed approach through a simulation example.

The problem of feature selection is a difficult combinatorial task in machine learning and of high practical relevance. In this paper, we consider feature selection method for multimodally distributed data, and present a large margin feature weighting method for k-nearest neighbor (kNN) classifiers. The method learns the feature weighting factors by minimizing a cost function, which aims at separating different classes by large local margins and pulling closer together points from the same class, based on using as few features as possible. The consequent optimization problem can be efficiently solved by linear programming. Finally, the proposed approach is assessed through a series of experiments with UCI and microarray data sets, as well as a more specific and challenging task, namely, radar high-resolution range profiles (HRRP) automatic target recognition (ATR). The experimental results demonstrate the effectiveness of the proposed algorithms.

Radar automatic target recognition (ATR) using high range resolution profiles (HRRPs) is addressed. A subband fusion structure is proposed based on wavelet packet transforming. Multiple adaptive Gaussian classifiers (AGCs) are built for each subband, and the outputs of each of the subband classifiers are combined to make a final decision. Compared with the traditional wideband recognition approach, i.e., single band approach, the proposed approach can achieve better recognition performance, and is more robust to noise as well. Example results, based on the measured data, show the efficiency of the proposed method.

In this paper, a heuristic algorithm based on a nonlinear nonconvex programming relaxation of the CDMA maximum likelihood (ML) problem is presented. Simulation results have shown that the BER performances of a detection strategy based on the heuristic algorithm are similar to that of the detection strategy based on the semidefinite relaxation. Furthermore, average CPU time of the heuristic algorithm is significantly lower than that of the randomized rounding algorithm based on a semidefinite relaxation. This approach provides good approximations to the ML performance.