Search

This paper presents a machine learning-based figure of merit model for superjunction (SJ) U-MOSFET (SSJ-UMOS) with a modulated drift region utilizing semi-insulating poly-crystalline silicon (SIPOS) pillars. This SJ drift region modulation is achieved through SIPOS pillars beneath the trench gate, focusing on optimizing the tradeoff between breakdown voltage (BV) and specific ON-resistance (RON,sp). This analytical model considers the effects of electric field modulation, charge-coupling, and majority carrier accumulation due to additional SIPOS pillars. Gaussian process regression is employed for the figure of merit (FOM = BV2/RON,sp) prediction and hyperparameter optimization, ensuring a reasonable and accurate model. A methodology is devised to determine the optimal BV-RON,sp tradeoff, surpassing the SJ silicon limit. The paper also delves into a discussion of optimal structural parameters for drift region, oxide thickness, and electric field modulation coefficients within the analytical model. The validity of the proposed model is robustly confirmed through comprehensive verification against TCAD simulation results. [ABSTRACT FROM AUTHOR]

Graph convolutional networks (GCNs) are neural network frameworks for machine learning on graphs. They can simultaneously perform end-to-end learning on the attribute information and the structure information of graph data. However, most existing GCNs inevitably encounter the limitations of non-robustness and low classification accuracy when labeled nodes are scarce. To address the two issues, the deep graph convolutional generative adversarial network (DGCGAN), a model combining GCN and deep convolutional generative adversarial networks (DCGAN), is proposed in this paper. First, the graph data is mapped to a highly nonlinear space by using the topology and attribute information of the graph for symmetric normalized Laplacian transform. Then, through the feature-structured enhanced module, the node features are expanded into regular structured data, such as images and sequences, which are input to DGCGAN as positive samples, thus expanding the sample capacity. In addition, the feature-enhanced (FE) module is adopted to enhance the typicality and discriminability of node features, and to obtain richer and more representative features, which is helpful for facilitating accurate classification. Finally, additional constraints are added to the network model by introducing DCGAN, thus enhancing the robustness of the model. Through extensive empirical studies on several standard benchmarks, we find that DGCGAN outperforms state-of-the-art baselines on semi-supervised node classification and remote sensing image classification. [ABSTRACT FROM AUTHOR]

The aim of remote sensing image captioning (RSIC) is to describe a given remote sensing image (RSI) using coherent sentences. Most existing attention-based methods model the coherence through an LSTM-based decoder, which dynamically infers a word vector from preceding sentences. However, these methods are indirectly guided through the confusion of attentive regions, as (1) the weighted average in the attention mechanism distracts the word vector from capturing pertinent visual regions and (2) there are few constraints or rewards for learning long-range transitions. In this paper, we propose a multi-source interactive stair attention mechanism that separately models the semantics of preceding sentences and visual regions of interest. Specifically, the multi-source interaction takes previous semantic vectors as queries and applies an attention mechanism on regional features to acquire the next word vector, which reduces immediate hesitation by considering linguistics. The stair attention divides the attentive weights into three levels—that is, the core region, the surrounding region, and other regions—and all regions in the search scope are focused on differently. Then, a CIDEr-based reward reinforcement learning is devised, in order to enhance the quality of the generated sentences. Comprehensive experiments on widely used benchmarks (i.e., the Sydney-Captions, UCM-Captions, and RSICD data sets) demonstrate the superiority of the proposed model over state-of-the-art models, in terms of its coherence, while maintaining high accuracy. [ABSTRACT FROM AUTHOR]

To learn discriminative features, hyperspectral image (HSI), containing 3-D cube data, is a preferable means of capturing multi-head self-attention from both spatial and spectral domains if the burden in model optimization and computation is low. In this paper, we design a dual multi-head contextual self-attention (DMuCA) network for HSI classification with the fewest possible parameters and lower computation costs. To effectively capture rich contextual dependencies from both domains, we decouple the spatial and spectral contextual attention into two sub-blocks, SaMCA and SeMCA, where depth-wise convolution is employed to contextualize the input keys in the pure dimension. Thereafter, multi-head local attentions are implemented as group processing when the keys are alternately concatenated with the queries. In particular, in the SeMCA block, we group the spatial pixels by evenly sampling and create multi-head channel attention on each sampling set, to reduce the number of the training parameters and avoid the storage increase. In addition, the static contextual keys are fused with the dynamic attentional features in each block to strengthen the capacity of the model in data representation. Finally, the decoupled sub-blocks are weighted and summed together for 3-D attention perception of HSI. The DMuCA module is then plugged into a ResNet to perform HSI classification. Extensive experiments demonstrate that our proposed DMuCA achieves excellent results over several state-of-the-art attention mechanisms with the same backbone. [ABSTRACT FROM AUTHOR]

Hyperspectral images (HSIs) have gained high spectral resolution due to recent advances in spectral imaging technologies. This incurs problems, such as an increased data scale and an increased number of bands for HSIs, which results in a complex correlation between different bands. In the applications of remote sensing and earth observation, ground objects represented by each HSI pixel are composed of physical and chemical non-Euclidean structures, and HSI classification (HIC) is becoming a more challenging task. To solve the above problems, we propose a framework based on a deep attention graph convolutional network (DAGCN). Specifically, we first integrate an attention mechanism into the spectral similarity measurement to aggregate similar spectra. Therefore, we propose a new similarity measurement method, i.e., the mixed measurement of a kernel spectral angle mapper and spectral information divergence (KSAM-SID), to aggregate similar spectra. Considering the non-Euclidean structural characteristics of HSIs, we design deep graph convolutional networks (DeepGCNs) as a feature extraction method to extract deep abstract features and explore the internal relationship between HSI data. Finally, we dynamically update the attention graph adjacency matrix to adapt to the changes in each feature graph. Experiments on three standard HSI data sets, namely, the Indian Pines, Pavia University, and Salinas data sets, demonstrate that the DAGCN outperforms the baselines in terms of various evaluation criteria. For example, on the Indian Pines data set, the overall accuracy of the proposed method achieves 98.61% when the training sample is 10%. [ABSTRACT FROM AUTHOR]

Anomaly detection of a hyperspectral image without any prior information has attracted much more attention in remote sensing image understanding and interpretation, which aims at determining whether a sample belongs to background or anomaly. Low-rank dictionary learning plays an important role in exploiting the low-rank prior of background for hyperspectral image (HSI) anomaly detection. In this article, the low-rank dictionary learning is introduced to learn a dictionary which can reconstruct the background positively, while anomaly cannot. Considering the high correlation of data especially between the adjacent bands, we resort to spectral-difference low-rank dictionary representation learning for global background modeling which can fully exploit the low-rank prior of background. Then, the residual matrix is used to distinguish anomaly. Different from the existing anomaly detection methods based on dictionary which is constructed or learned in a separated step, our proposed model can simultaneously learn the dictionary and separate anomaly by iterative learning. The experimental results on five real data sets demonstrate the superior performance of the proposed method for hyperspectral anomaly detection compared with other state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Recently, many stochastic variance reduced alternating direction methods of multipliers (ADMMs) (e.g., SAG-ADMM and SVRG-ADMM) have made exciting progress such as linear convergence rate for strongly convex (SC) problems. However, their best-known convergence rate for non-strongly convex (non-SC) problems is $\mathcal {O}(1/T)$ O (1 / T) as opposed to $\mathcal {O}(1/T^2)$ O (1 / T 2) of accelerated deterministic algorithms, where $T$ T is the number of iterations. Thus, there remains a gap in the convergence rates of existing stochastic ADMM and deterministic algorithms. To bridge this gap, we introduce a new momentum acceleration trick into stochastic variance reduced ADMM, and propose a novel accelerated SVRG-ADMM method (called ASVRG-ADMM) for the machine learning problems with the constraint $Ax + By = c$ A x + B y = c . Then we design a linearized proximal update rule and a simple proximal one for the two classes of ADMM-style problems with $B = \tau I$ B = τ I and $B\ne \tau I$ B ≠ τ I , respectively, where $I$ I is an identity matrix and $\tau$ τ is an arbitrary bounded constant. Note that our linearized proximal update rule can avoid solving sub-problems iteratively. Moreover, we prove that ASVRG-ADMM converges linearly for SC problems. In particular, ASVRG-ADMM improves the convergence rate from $\mathcal {O}(1/T)$ O (1 / T) to $\mathcal {O}(1/T^2)$ O (1 / T 2) for non-SC problems. Finally, we apply ASVRG-ADMM to various machine learning problems, e.g., graph-guided fused Lasso, graph-guided logistic regression, graph-guided SVM, generalized graph-guided fused Lasso and multi-task learning, and show that ASVRG-ADMM consistently converges faster than the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Deep learning (DL) has become a hot topic in the research field of hyperspectral image (HSI) classification. However, with increasing depth and size of deep learning methods, its application in mobile and embedded vision applications has brought great challenges. In this article, we address a network architecture search (NAS)-guided lightweight spectral–spatial attention feature fusion network (LMAFN) for HSI classification. The overall architecture of the proposed network is guided by several conclusions of NAS, which achieves fewer parameters and lower computation cost with deeper network structure by exploiting multiscale Ghost grouped with efficient channel attention (ECA) module for adaptively adjusting the weights of different channels. It helps fully extract spectral–spatial discriminant features to avoid information loss of the dimension reduction operation. Specifically, a multilayer feature fusion method is proposed to extract the fusion information of the spectral–spatial features of each layer by considering complementary information of different hierarchical structures. Therefore, high-lever spectral–spatial attributes are gradually exploited along with the increase in layers and the fusion of layers. The experimental verification on three real HSI data sets demonstrates that the proposed framework presents more satisfying classification performance and efficiency with deeper network structure and lower parameter size. [ABSTRACT FROM AUTHOR]

As a fundamental but challenging task in the interpretation of remote sensing images, scene classification plays an important role in various applications and has become an active research topic. Many previous works have demonstrated the remarkable performance of the deep convolutional neural networks (CNNs) for remote sensing scene classification. However, the progress made by CNN-based methods for scene classification has gradually reached saturation in recent years, due to the serious dependence on the pretrained CNN models, the limitations of manually designed network architecture and the disadvantages of existing data sets. In this article, a new paradigm to automatically design a suitable CNN architecture for scene classification is investigated. We propose an efficient architecture search framework to discover optimal network architectures in continuous search space with the gradient-based optimization method. Our framework consists of two stages: the search phase and the evaluation phase. During the search process, a greedy and progressive search strategy is introduced to search network building blocks (i.e., cells) through bilevel optimization. Besides, we propose a simple architecture regularization scheme to further improve the search efficiency and the robustness of discovered architectures. After the search process, the optimal cell architectures are determined and then repeatedly stacked to construct the final network for evaluation. For the data set, we propose a mergence strategy to build a new large-scale remote sensing scene image data set that contains rich scene categories and image diversity, making it feasible to find a new CNN model with strong generalization ability for scene classification. Extensive experiments demonstrate the efficiency of the proposed search strategies and the impressive classification performance of searched CNN architectures on seven public benchmark data sets, including four large-scale data sets and three small-scale data sets. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have demo- nstrated outstanding performance on image classification. To classify the hyperspectral images (HSIs), existing CNN-based approaches commonly adopt the architecture using single or several fixed spatial windows as inputs. This kind of architecture may lose contextual information or incorporate heterogeneous information due to the neglect of various land-cover distributions in HSIs. To deal with this problem, a novel attention multibranch CNN method based on adaptive region search (RS-AMCNN) is proposed for HSI classification. In RS-AMCNN, sizes and locations of spatial windows are searched in the nonlocal candidate region adaptively according to sample-specific distribution. These flexible spatial windows are input into several branches of RS-AMCNN. In each branch, convolutional long short-term memories (ConvLSTMs) are merged into CNN from shallow to deep layers, which not only extracts joint spatial–spectral features, but also exploits complementary information among different layers. Then, a branch attention mechanism is devised to emphasize more discriminative branches and suppress less useful ones. It forces RS-AMCNN to extract multiscale and multicontextual attention features for classification. Finally, RS-AMCNN is optimized end-to-end by combining the losses from the ramose classifiers of different branches and the main classifier. Experiments carried on several benchmark HSI data sets demonstrate that RS-AMCNN provides promising classification performance, especially in edge preservation and region uniformity. [ABSTRACT FROM AUTHOR]

In the last five years, deep learning has been introduced to tackle the hyperspectral image (HSI) classification and demonstrated good performance. In particular, the convolutional neural network (CNN)-based methods for HSI classification have made great progress. However, due to the high dimensionality of HSI and equal treatment of all bands, the performance of these methods is hampered by learning features from useless bands for classification. Moreover, for patchwise-based CNN models, equal treatment of spatial information from the pixel-centered neighborhood also hinders the performance of these methods. In this article, we propose an end-to-end residual spectral–spatial attention network (RSSAN) for HSI classification. The RSSAN takes raw 3-D cubes as input data without additional feature engineering. First, a spectral attention module is designed for spectral band selection from raw input data by emphasizing useful bands for classification and suppressing useless bands. Then, a spatial attention module is designed for the adaptive selection of spatial information by emphasizing pixels from the same class as the center pixel or those are useful for classification in the pixel-centered neighborhood and suppressing those from a different class or useless. Second, two attention modules are also used in the following CNN for adaptive feature refinement in spectral–spatial feature learning. Third, a sequential spectral–spatial attention module is embedded into a residual block to avoid overfitting and accelerate the training of the proposed model. Experimental studies demonstrate that the RSSAN achieved superior classification accuracy compared with the state of the art on three HSI data sets: Indian Pines (IN), University of Pavia (UP), and Kennedy Space Center (KSC). [ABSTRACT FROM AUTHOR]

Deep learning techniques represented by deep convolutional neural networks (CNNs) have been widely used in polarimetric synthetic-aperture radar (PolSAR) image classification in recent years. One challenge is how to get a pleasant classification result with limited human-labeled samples. In this letter, a novel semisupervised classification method incorporating labels’ semantic priors is proposed for PolSAR image classification with limited labeled samples. The core idea is that a good classification result map should have consistent regions and aligned boundaries. Thus, a cost function is proposed, which mainly contains three terms including a supervised term, a region consistency term, and a boundary kept term. The supervised term enforces the category label to be the same with the human-labeled labels. The region consistency term encourages the labels in one region to be consistent. The boundary-kept term constrains the region consistency term, preventing the classification map from being too smooth. An alternate iterative optimization method is proposed to solve this equation. First, a CNN is trained using the labeled samples and the classification probability map. Then, the classification probability map is updated by the prediction of the trained CNN and the labeled samples. Repeat these two procedures until the maximum number of iterations is met. Experiments on two real PolSAR images are conducted to validate the effectiveness of the proposed method compared with several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Polarimetric synthetic aperture radar (PolSAR) image classification plays an important role in remote sensing image processing. In recent years, stacked auto-encoder (SAE) has obtained a series of excellent results in PolSAR image classification. The recently proposed projective dictionary pair learning (DPL) model takes both accuracy and time consumption into consideration, and another recently proposed semicoupled dictionary learning (SCDL) model gives a new way to fit different features. Based on the SAE, DPL, and SCDL models, we propose a novel semicoupled projective DPL method with SAE (SAE-SDPL) for PolSAR image classification. Our method can get the classification result efficiently and correctly and meanwhile giving a new method to fit different features. In this paper, three PolSAR images are used to test the performance of SAE-SDPL. Compared with some state-of-the-art methods, our method obtains excellent results in PolSAR image classification. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new anomaly detection method for hyperspectral images based on two well-designed dictionaries: background dictionary and potential anomaly dictionary. In order to effectively detect an anomaly and eliminate the influence of noise, the original image is decomposed into three components: background, anomalies, and noise. In this way, the anomaly detection task is regarded as a problem of matrix decomposition. Considering the homogeneity of background and the sparsity of anomalies, the low-rank and sparse constraints are imposed in our model. Then, the background and potential anomaly dictionaries are constructed using the background and anomaly priors. For the background dictionary, a joint sparse representation (JSR)-based dictionary selection strategy is proposed, assuming that the frequently used atoms in the overcomplete dictionary tend to be the background. In order to make full use of the prior information of anomalies hidden in the scene, the potential anomaly dictionary is constructed. We define a criterion, i.e., the anomalous level of a pixel, by using the residual calculated in the JSR model within its local region. Then, it is combined with a weighted term to alleviate the influence of noise and background. Experiments show that our proposed anomaly detection method based on potential anomaly and background dictionaries construction can achieve superior results compared with other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new feature selection method called kernel fisher discriminant analysis and regression learning based algorithm for unsupervised feature selection. The existing feature selection methods are based on either manifold learning or discriminative techniques, each of which has some shortcomings. Although some studies show the advantages of two-steps method benefiting from both manifold learning and discriminative techniques, a joint formulation has been shown to be more efficient. To do so, we construct a global discriminant objective term of a clustering framework based on the kernel method. We add another term of regression learning into the objective function, which can impose the optimization to select a low-dimensional representation of the original dataset. We use L2,1-norm of the features to impose a sparse structure upon features, which can result in more discriminative features. We propose an algorithm to solve the optimization problem introduced in this paper. We further discuss convergence, parameter sensitivity, computational complexity, as well as the clustering and classification accuracy of the proposed algorithm. In order to demonstrate the effectiveness of the proposed algorithm, we perform a set of experiments with different available datasets. The results obtained by the proposed algorithm are compared against the state-of-the-art algorithms. These results show that our method outperforms the existing state-of-the-art methods in many cases on different datasets, but the improved performance comes with the cost of increased time complexity. [ABSTRACT FROM AUTHOR]

Existing polarimetric synthetic aperture radar (PolSAR) image classification methods cannot achieve satisfactory performance on complex scenes characterized by several types of land cover with significant levels of noise or similar scattering properties across land cover types. Hence, we propose a supervised classification method aimed at constructing a classifier based on self-paced learning (SPL). SPL has been demonstrated to be effective at dealing with complex data while providing classifier performance improvement. In this paper, a novel support vector machine (SVM) algorithm based on SPL with neighborhood constraints (SVM_SPLNC) is proposed. The proposed method leverages the easiest samples first to obtain an initial parameter vector. Then, more complex samples are gradually incorporated to update the parameter vector iteratively. Moreover, neighborhood constraints are introduced during the training process to further improve performance. Experimental results on three real PolSAR images show that the proposed method performs well on complex scenes. [ABSTRACT FROM AUTHOR]

In hyperspectral image processing, classification is one of the most popular research topics. In recent years, research progress made in deep-learning-based hierarchical feature extraction and classification has shown a great power in many applications. In this paper, we propose a novel local spatial sequential (LSS) method, which is used in a recurrent neural network (RNN). Using this model, we can extract local and semantic information for hyperspectral image classification. First, we extract low-level features from hyperspectral images, including texture and differential morphological profiles. Second, we combine the low-level features together and propose a method to construct the LSS features. Afterwards, we build an RNN and use the LSS features as the input to train the network for optimizing the system parameters. Finally, the high-level semantic features generated by the RNN is fed into a softmax layer for the final classification. In addition, a nonlocal spatial sequential method is presented for the recurrent neural network model (NLSS-RNN) to further enhance the classification performance. NLSS-RNN finds nonlocal similar structures to a given pixel and extracts corresponding LSS features, which not only preserve the local spatial information, but also integrate the information of nonlocal similar samples. The experimental results on three publicly accessible datasets show that our proposed method can obtain competitive performance compared with several state-of-the-art classifiers. [ABSTRACT FROM AUTHOR]

Deep learning has obtained state-of-the-art results in a variety of computer vision tasks and has also been used to solve SAR image classification problems. Deep learning algorithms typically require a large amount of training data to achieve high accuracy. In contrast, the size of SAR image datasets is often comparatively limited. Therefore, this paper proposes a novel method, deep memory convolution neural networks (M-Net), to alleviate the problem of overfitting caused by insufficient SAR image samples. Based on the convolutional neural networks (CNN), M-Net adds an information recorder to remember and store samples’ spatial features, and then it uses spatial similarity information of the recorded features to predict unknown sample labels. M-Net's use of this information recorder may cause difficulties for convergence if conventional CNN training methods were directly used to train M-Net. To overcome this problem, we propose a transfer parameter technique to train M-Net in two steps. The first step is to train a CNN, which has the same structure as the part of CNN incorporated in M-Net, to obtain initial training parameters. The second step applies the initialized parameters to M-Net and then trains the entire M-Net. This two-step training approach helps us to overcome the nonconvergence issue, and also reduces training time. We evaluate M-Net using the public benchmark MSTAR dataset, and achieve higher accuracy than several other well-known SAR image classification algorithms. [ABSTRACT FROM AUTHOR]

The receiver operating characteristic (ROC) and detection error tradeoff (DET) curves have been widely used in the machine learning community to analyze the performance of classifiers. The area (or volume) under the convex hull has been used as a scalar indicator for the performance of a set of classifiers in ROC and DET space. Recently, 3D convex-hull-based evolutionary multiobjective optimization algorithm (3DCH-EMOA) has been proposed to maximize the volume of convex hull for binary classification combined with parsimony and three-way classification problems. However, 3DCH-EMOA revealed high consumption of computational resources due to redundant convex hull calculations and a frequent execution of nondominated sorting. In this paper, we introduce incremental convex hull calculation and a fast replacement for non-dominated sorting. While achieving the same high quality results, the computational effort of 3DCH-EMOA can be reduced by orders of magnitude. The average time complexity of 3DCH-EMOA in each generation is reduced from O ( n 2 log n ) to O ( n log n ) per iteration, where n is the population size. Six test function problems are used to test the performance of the newly proposed method, and the algorithms are compared to several state-of-the-art algorithms, including NSGA-III, RVEA, etc., which were not compared to 3DCH-EMOA before. Experimental results show that the new version of the algorithm (3DFCH-EMOA) can speed up 3DCH-EMOA for about 30 times for a typical population size of 300 without reducing the performance of the method. Besides, the proposed algorithm is applied for neural networks pruning, and several UCI datasets are used to test the performance. [ABSTRACT FROM AUTHOR]

Under the umbrella of tensor algebra, this paper proposes a new sparse-coding-based classifier (SCC) for hyperspectral imagery classification (HIC). By utilizing the tensor forms of hyperspectral pixels, we advance a tensor sparse-coding model which preserves as many original spatial constraints of a pixel and its spatial neighbors as possible. Furthermore, to alleviate the classification uncertainty resulted from widely existing mixed pixels, this paper constructs a regularization term for maximizing the likelihood of sparse-coding tensor defined on the posterior class probability. By combining the tensor sparse coding with maximizing likelihood estimation, a hybrid probabilistic SCC with spatial neighbor tensor (HPSCC-SNT) is proposed, which makes the pixels be well represented by the training pixels belonging to the same class. The performance of HPSCC-SNT is evaluated on three real hyperspectral imagery data sets, and the results show that it can achieve accurate and robust HIC results, and outperforms the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Sparsity-regularized synthetic aperture radar (SAR) imaging framework has shown its remarkable performance to generate a feature-enhanced high-resolution image, in which a sparsity-inducing regularizer is involved by exploiting the sparsity priors of some visual features in the underlying image. However, since the simple prior of low-level features is insufficient to describe different semantic contents in the image, this type of regularizer will be incapable of distinguishing between the target of interest and unconcerned background clutters. As a consequence, the features belonging to the target and clutters are simultaneously affected in the generated image without concerning their underlying semantic labels. To address this problem, we propose a novel semantic information guided generative framework for target-oriented SAR image formation, which aims at enhancing the interested target scatters while suppressing the background clutters. First, we develop a new semantics-specific regularizer for image formation by exploiting the statistical properties of different semantic categories in a target scene SAR image. In order to infer the semantic label for each pixel in an unsupervised way, we moreover induce a novel high-level prior-driven regularizer and some semantic causal rules from the prior knowledge. Finally, our regularized framework for image formation is further derived as a simple iteratively reweighted $\ell _{1}$ minimization problem that can be conveniently solved by many off-the-shelf solvers. Experimental results demonstrate the effectiveness and superiority of our framework for SAR image formation in terms of target enhancement and clutters suppression, compared with the state of the arts. Additionally, the proposed framework opens a new direction of devoting some machine learning strategies to image formation, which can benefit the subsequent decision-making tasks. [ABSTRACT FROM AUTHOR]

Semi-supervised learning is one of the most promising learning paradigms in many practical applications where few labeled samples are available. Among such learning models, semi-supervised support vector machine (SVM) is a typical and prominent one. However, the typical semi-supervised SVM cannot avoid estimating the distribution of positive and negative samples. In this paper, we put forward a method with the name of EnsembleS3VM which deals with the unknown distribution by ensemble learning. It builds a semi-supervised SVM model composed of base learners based on different disturbance factors, and raises an ensemble method based on clustering evaluation means. Meanwhile, it presents a combination of two multi-classification strategies in order to reduce the running time and enhance the classification accuracy simultaneously. The proposed method can deal with semi-supervised classification problems even with unknown distribution or unbalanced data. Experiments on UCI datasets prove the effectiveness of ensemble strategy and the robustness under different sample distributions. We also apply the proposed algorithm to a practical application, i.e., ground cover classification for polarimetric synthetic aperture radar images which is a typical but difficult semi-supervised classification problem. [ABSTRACT FROM AUTHOR]

Wisely utilizing the internal and external learning methods is a new challenge in super-resolution problem. To address this issue, we analyze the attributes of two methodologies and find two observations of their recovered details: 1) they are complementary in both feature space and image plane and 2) they distribute sparsely in the spatial space. These inspire us to propose a low-rank solution which effectively integrates two learning methods and then achieves a superior result. To fit this solution, the internal learning method and the external learning method are tailored to produce multiple preliminary results. Our theoretical analysis and experiment prove that the proposed low-rank solution does not require massive inputs to guarantee the performance, and thereby simplifying the design of two learning methods for the solution. Intensive experiments show the proposed solution improves the single learning method in both qualitative and quantitative assessments. Surprisingly, it shows more superior capability on noisy images and outperforms state-of-the-art methods. [ABSTRACT FROM PUBLISHER]

In this paper, a novel polarimetric synthetic aperture radar (PolSAR) image classification method based on multilayer autoencoders and self-paced learning (SPL) is proposed. The multilayer autoencoders network is used to learn the features, which convert raw data into more abstract expressions. Then, softmax regression is applied to produce the predicted probability distributions over all the classes of each pixel. When we optimize the multilayer autoencoders network, self-paced learning is used to accelerate the learning convergence and achieve a stronger generalization capability. Under this learning paradigm, the network learns the easier samples first and gradually involves more difficult samples in the training process. The proposed method achieves the overall classification accuracies of 94.73%, 94.82% and 78.12% on the Flevoland dataset from AIRSAR, Flevoland dataset from RADARSAT-2 and Yellow River delta dataset, respectively. Such results are comparable with other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Community structure has become one of the central studies of the topological structure of complex networks in the past decades. Although many advanced approaches have been proposed to identify community structure, those state-of-the-art methods still lack efficiency in terms of a balance between stability, accuracy and computation time. Here, we propose an algorithm with different stages, called TJA-net, to efficiently identify communities in a large network with a good balance between accuracy, stability and computation time. First, we propose an initial labeling algorithm, called ILPA, combining K-nearest neighbor (KNN) and label propagation algorithm (LPA). To produce a number of sub-communities automatically, ILPA iteratively labels a node in a network using the labels of its adjacent nodes and their index of closeness. Next, we merge sub-communities using the mutual membership of two communities. Finally, a refinement strategy is designed for modifying the label of the wrongly clustered nodes at boundaries. In our approach, we propose and use modularity density as the objective function rather than the commonly used modularity. This can deal with the issue of the resolution limit for different network structures enhancing the result precision. We present a series of experiments with artificial and real data set and compare the results obtained by our proposed algorithm with the ones obtained by the state-of-the-art algorithms, which shows the effectiveness of our proposed approach. The experimental results on large-scale artificial networks and real networks illustrate the superiority of our algorithm. [ABSTRACT FROM AUTHOR]

For hyperspectral image (HSI) classification, it is very important to learn effective features for the discrimination purpose. Meanwhile, the ability to combine spectral and spatial information together in a deep level is also important for feature learning. In this letter, we propose an unsupervised feature learning method for HSI classification, which is based on recursive autoencoders (RAE) network. RAE utilizes the spatial and spectral information and produces high-level features from the original data. It learns features from the neighborhood of the investigated pixel to represent the whole local homogeneous area of the image. In addition, to obtain more accurate representation of the investigated pixel, a weighting scheme is adopted based on the neighboring pixels, where the weights are determined by the spectral similarity between the neighboring pixels and the investigated pixel. The effectiveness of our method is evaluated by the experiments on two hyperspectral data sets, and the results show that our proposed method has a better performance. [ABSTRACT FROM PUBLISHER]

The past decade has witnessed an explosive growth of data streams in Internet, biometrics, remote sensing and other fields. Nowadays many supervised incremental/online learning approaches have been developed, to avoid retraining and reduce the computational complexity when data come chunk by chunk. However, these methods can’t obtain satisfied performance when the labeled samples are limited. In this paper, we propose an Incremental Laplacian Regularization Extreme Learning Machine (ILR-ELM) for semi-supervised online learning, by utilizing both labeled and unlabeled samples. Unlike most of the existing semi-supervised incremental/online learning algorithms, this paper not only proposes incremental/online learning mechanism for data chunk containing both labeled and unlabeled samples but also proposes incremental/online learning mechanism for data chunk containing only unlabeled samples. The latter case is more common in practical applications because there is usually no enough time to label the samples for continuously arriving data stream. The alternative analytical solutions of ILR-ELM for the two incremental/online learning mechanisms are also presented. The performance of ILR-ELM is evaluated on three benchmark machine learning data, and the results show that it can achieve near accurate and robust classification/regression with a small number of labeled data, and outperforms the incremental/online learning approaches. Compared with the supervised and the comparative semi-supervised incremental/online learning methods, the generalization accuracy of ILR-ELM is increased by nearly 9% and 2% respectively for the classification problem. For the regression problem, the generalization error of ILR-ELM is reduced by nearly 14% than the supervised incremental/online learning methods and 1% than compared semi-supervised incremental/online learning methods Furthermore, ILR-ELM can achieve comparable prediction to semi-supervised batch learning, using less time and Random Access Memory (RAM). [ABSTRACT FROM AUTHOR]

Recently, very high resolution (VHR) panchromatic and multispectral (MS) remote-sensing images can be acquired easily. However, it is still a challenging task to fuse and classify these VHR images. Generally, there are two ways for the fusion and classification of panchromatic and MS images. One way is to use a panchromatic image to sharpen an MS image, and then classify a pan-sharpened MS image. Another way is to extract features from panchromatic and MS images, respectively, and then combine these features for classification. In this paper, we propose a superpixel-based multiple local convolution neural network (SML-CNN) model for panchromatic and MS images classification. In order to reduce the amount of input data for the CNN, we extend simple linear iterative clustering algorithm for segmenting MS images and generating superpixels. Superpixels are taken as the basic analysis unit instead of pixels. To make full advantage of the spatial-spectral and environment information of superpixels, a superpixel-based multiple local regions joint representation method is proposed. Then, an SML-CNN model is established to extract an efficient joint feature representation. A softmax layer is used to classify these features learned by multiple local CNN into different categories. Finally, in order to eliminate the adverse effects on the classification results within and between superpixels, we propose a multi-information modification strategy that combines the detailed information and semantic information to improve the classification performance. Experiments on the classification of Vancouver and Xi’an panchromatic and MS image data sets have demonstrated the effectiveness of the proposed approach. [ABSTRACT FROM PUBLISHER]

Discrete multiple signal classification (MUSIC) with its low computational cost and mild condition requirement becomes a significant noniterative algorithm for joint sparse recovery (JSR). However, it fails in rank defective problem caused by coherent or limited amount of multiple measurement vectors (MMVs). In this letter, we provide a novel sight to address this problem by interpreting JSR as a binary classification problem with respect to atoms. Meanwhile, MUSIC essentially constructs a supervised classifier based on the labeled MMVs so that its performance will heavily depend on the quality and quantity of these training samples. From this viewpoint, we develop a semisupervised MUSIC (SS-MUSIC) in the spirit of machine learning, which declares that the insufficient supervised information in the training samples can be compensated from those unlabeled atoms. Instead of constructing a classifier in a fully supervised manner, we iteratively refine a semisupervised classifier by exploiting the labeled MMVs and some reliable unlabeled atoms simultaneously. Through this way, the required conditions and iterations can be greatly relaxed and reduced. Numerical experimental results demonstrate that SS-MUSIC can achieve much better recovery performances than other MUSIC extended algorithms as well as some typical greedy algorithms for JSR in terms of iterations and recovery probability. [ABSTRACT FROM PUBLISHER]

As an advanced nonlinear technique, deep learning, which is based on deep neural networks (DNNs), has attracted considerable attentions. In this paper, we propose a novel neighborhood preserved deep neural network (NPDNN) for polarimetric synthetic aperture radar feature extraction and classification. The spatial relation between pixels is exploited by a jointly weighting strategy. Not only the spatial neighbors but also the pixels in the same superpixel are utilized to weight each pixel. This strategy maintains the spatial dependence leading to superior homogeneity of the terrains without extra computational memory. Moreover, a few labeled samples and their nearest neighbors are employed to train the multilayer NPDNN, which preserves the local structure and reduces the number of labeled samples for classification. Experimental results on synthesized and real PolSAR data show that the proposed NPDNN can improve the classification accuracy compared with state-of-the-art DNNs despite a few input samples. [ABSTRACT FROM PUBLISHER]

Clustering learning and classification learning are two major tasks in pattern recognition. The traditional hybrid clustering and classification algorithms handle them in a sequential way rather than a simultaneous way. Fortunately, multiobjective optimization provides a way to solve this problem. In this paper, an algorithm that learns simultaneous clustering and classification adaptively via multiobjective evolutionary algorithm is proposed. The main idea of this paper is to optimize two objective functions which represent fuzzy cluster connectedness and classification error rate to achieve the goal of simultaneous learning. Firstly, we adopt a graph based representation scheme to encode so that it can generate a set of solutions with different number of clusters in a single run. Then the relationship between clustering and classification is built via the Bayesian theory during the optimization process. The quality of clustering and classification is measured by the objective functions and the feedback drawn from both aspects is used to guide the mutation. At last, a set of nondominated solutions are generated, from which the final Pareto optimal solution is selected by using Adjusted Rand Index. The results on synthetic datasets and real-life datasets demonstrate the rationality and effectiveness of the proposed algorithm. Furthermore, we apply the proposed algorithm to image segmentation including texture images and synthetic aperture radar images, the experimental results show the superiority of the proposed algorithm compared with other five algorithms. [ABSTRACT FROM AUTHOR]

Incremental learning has been developed for supervised classification, where knowledge is accumulated incrementally and represented in the learning process. However, labeling sufficient samples in each data chunk is of high cost, and incremental technologies are seldom discussed in the semi-supervised paradigm. In this paper we advance an Incremental Semi-Supervised classification approach via Self-Representative Selection (IS 3 RS) for data streams classification, by exploring both the labeled and unlabeled dynamic samples. An incremental self-representative data selection strategy is proposed to find the most representative exemplars from the sequential data chunk. These exemplars are incrementally labeled to expand the training set, and accumulate knowledge over time to benefit future prediction. Extensive experimental evaluations on some benchmarks have demonstrated the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]

It remains one of the most challenging tasks to distinguish different terrain materials from a single SAR image. With the increase of ground resolution, it allows us to model the SAR image directly by exploiting spatial structures and texture information that are extracted by several machine learning approaches. In this paper, a novel feature learning approach is proposed to capture discriminant features of high-resolution SAR images. In the first stage, a weighted discriminant filter bank is learned from some labeled SAR image patches to generate low-level features. Then, the locality constraint is introduced to produce the high-level features in both the encoding and the spatial pooling procedure. In this work, the superpixels are employed as the basic operational units instead of the pixels for terrain classification. With some learned domain patterns which are learned from all of the high-level features of each pixel, the superpixel is characterized by a hyper-feature. In the last stage, a linear-kernel support vector machine is utilized to classify all of these hyper-features which are generated for each superpixel. The experimental results show a better classification performance of the proposed approach than several available state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Time series classification (TSC) is a crucial and challenging problem in sequential analysis. However, most of the existing best-performing methods are time-consuming, even if coping with small-scale datasets. Broad learning systems (BLS) have shown low time complexity and high accuracy in handling various tasks and have been applied to many fields. However, none of the BLS-based methods is suitable to tackle TSC due to (1) unsuitable structure in capturing temporal information of time series; (2) low interpretability in representing the evolving relations among states along time. Thus, this paper develops a broad fuzzy cognitive map system (BFCMS) to address time series classification efficiently, which consists of the sparse autoencoder (SAE) based feature extraction block, the high-order fuzzy cognitive map (HFCM) based spatiotemporal information aggregation block, and one multilayer perceptron (MLP) based prediction layer. The feature extraction block is designed to capture the underlying core evolving patterns, and the spatiotemporal information aggregation block is developed to model the underlying causal relationships and contextual dependencies. These two blocks are designed to overcome the limitations of BLS. MLP is applied to map the feature representation to the label of time series based on the aggregated feature representations from these two blocks. In addition, BFCMS develops three incremental learning strategies for fast updating in broad expansion without a retraining procedure if the model deems to be expanded. We compared BFCMS with other state-of-the-art baselines on 26 datasets. The experimental results demonstrate the superiority of BFCMS. Concretely, BFCMS achieves a lower training cost with on-par classification accuracy. • BFCMS can balance the time complexity and accuracy of the time series classifier. • We design the feature extraction and spatiotemporal information aggregation blocks to help BLS handle TSC tasks. • BFCMS retains the advantages of BLS in flexibility. [ABSTRACT FROM AUTHOR]

The deep learning algorithm has made great breakthroughs in optical image processing. Some deep learning algorithms require a large number of labeled samples for training. For PolSAR data sets, due to the influence of speckle noise and other factors, high-quality labeled data are limited. Therefore, it is meaningful to use deep learning algorithm to solve PolSAR classification problem in limited labeled dataset. This paper proposes a spatial feature-based convolutional neural network (SF-CNN). The network adopts a dual-branch CNN structure. Both of the two branches have the same structure and share parameters. SF-CNN can receive more than one sample as input. SF-CNN's special structure can expand the original training set by combining different samples, and alleviate the problem of insufficient labeled training data in PolSAR image classification tasks. When training, SF-CNN maps high-dimensional PolSAR image to low-dimensional feature space. In low-dimensional feature space, SF-CNN enhances the ability of network to extract discriminative features by maximizing or minimizing the distance between feature centers of different classes. In order to dig up the relationship between the samples, the test sample features are compared with every training sample feature when testing. Finally, labels of test samples are determined by the comparison result. The result of SF-CNN in PolSAR image classification task is better than that of standard CNN. • This paper proposes a spatial feature-based convolutional neural network (SF-CNN). • SF-CNN adopts a dual-branch CNN structure. • Both of the two branches have the same structure and share parameters. • SF-CNN can receive more than one sample as input. • SF-CNN's special structure can expand the original training set. [ABSTRACT FROM AUTHOR]

Dimension reduction has always been a major problem in many applications of machine learning and pattern recognition. In this paper, the scaling cut criterion-based supervised dimension reduction methods for data analysis are proposed. The scaling cut criterion can eliminate the limit of the hypothesis that data distribution of each class is homoscedastic Gaussian. To obtain a more reasonable mapping matrix and reduce the computational complexity, local scaling cut criterion-based dimension reduction is raised, which utilized the localization strategy of the input data. The localized $$k$$ -nearest neighbor graph is introduced , which relaxes the within-class variance and enlarges the between-class margin. Moreover, by kernelizing the scaling cut criterion and local scaling cut criterion, both methods are extended to efficiently model the nonlinear variability of the data. Furthermore, the optimal dimension scaling cut criterion is proposed, which can automatically select the optimal dimension for the dimension reduction methods. The approaches have been tested on several datasets, and the results have shown a better and efficient performance compared with other linear and nonlinear dimension reduction techniques. [ABSTRACT FROM AUTHOR]

Ensemble learning has been a hot topic in machine learning due to its successful utilization in many applications. Rough set theory has been proved to be an excellent mathematical tool for dimension reduction. In this paper, based on rough set, a novel framework for ensemble is proposed. In our proposed framework, the relationship among attributes in rough subspace is first considered, and the maximum dependency degree of attribute is first employed to effectively reduce the searching space of reducts and augment the diversity of selected reducts. In addition, in order to choose an appropriate reduct from the dynamic reduct searching space, an assessment function which can balance the accuracy and diversity is utilized. At last, a new method, i.e., Dynamic Rough Subspace based Selective Ensemble (DRSSE), which is derived from our framework is given. By repeatedly changing the searching space of reducts and selecting the next reduct from the changed searching space, DRSSE finally trains an ensemble system with these selected reducts. Compared with several available ensemble methods, experimental results with several datasets demonstrate that DRSSE can lead to a comparative or even better performance. [ABSTRACT FROM AUTHOR]

Diversity and accuracy are the two key factors that decide the ensemble generalization error. Constructing a good ensemble method by balancing these two factors is difficult, because increasing diversity is at the cost of reducing accuracy normally. In order to improve the performance of an ensemble while avoiding the difficulty derived of balancing diversity and accuracy, we propose a novel method that weights each classifier in the ensemble by maximizing three different quadratic forms. In this paper, the optimal weight of individual classifiers is obtained by minimizing the ensemble error, rather than analyzing diversity and accuracy. Since it is difficult to minimize the general form of the ensemble error directly, we approximate the error in an objective function subject to two constraints ( ∑ w i = 1 and − 1 < w i < 1 ). Particularly, we introduce an error term with a weight vector w 0 , and subtract this error with the quadratic form to obtain our approximated error. This subtraction makes minimizing the approximation form equivalent to maximizing the original quadratic form. Theoretical analysis finds that when the value of the quadratic form is maximized, the error of an ensemble system with the corresponding optimal weight w * will be smallest, especially compared with the ensemble with w 0 . Finally, we demonstrate improved classification performance from the experimental results of an artificial dataset, UCI datasets and PolSAR image data. [ABSTRACT FROM AUTHOR]

We propose adaptive constraint propagation (ACP) for semi-supervised kernel matrix learning (SS-KML). SS-KML aims to learn a kernel matrix from the given samples which contains just a little supervised information such as class labels or pairwise constraints. Recently, for effective SS-KML, constraint propagation methods by semi-definite programming are being actively studied, and the representative works are pairwise constraint propagation (PCP) and kernel propagation (KP). They have used hard constraints in their frameworks for constraint propagation and achieved outstanding classification performance. However, a small set of the hard constraints sometimes cannot cover all the samples of the full data set, and thus lead to large distortions of global discriminative data structure in the learned kernel matrix. It has a negative influence on the classification performance. To deal with this problem, we provide two adaptive fidelity terms to satisfy the requirement that two must-link samples become close and two cannot-link samples become far apart. Then, we build a new framework based on them to adaptively propagate the constraints. Experimental results demonstrate that ACP outperforms state-of-the-art SS-KML methods such as PCP and KP in terms of both effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

The learning vector quantization (LVQ) algorithm is widely used in image compression because of its intuitively clear learning process and simple implementation. However, LVQ strongly depends on the initialization of the codebook and often converges to local optimal results. To address the issues, a new two-step LVQ (TsLVQ) algorithm is proposed in the paper. TsLVQ uses a correcting learning stage after LVQ to move the synaptic weight vector away from the incorrectly clustered training vector and towards the correctly clustered training vector. Experimental results show that TsLVQ outperforms kernel-based LVQ (KLVQ) and LVQ in terms of peak signal-to-noise ratio. [ABSTRACT FROM PUBLISHER]

In the increasingly complex electromagnetic environment, a variety of new signal types are appearing; however, existing electromagnetic signal classification (ESC) models cannot handle new signal types. In this context, the emergence of class-incremental learning aims to incrementally update the classification model as new categories emerge. In this paper, an electromagnetic signal classification framework based on class exemplar selection and a multi-objective linear programming classifier (CES-MOLPC) is proposed in order to continuously learn new classes in an incremental manner. Specifically, our approach involves the adaptive selection of class exemplars considering normalized mutual information and a multi-objective linear programming classifier. The former is used to maintain the classification capability of the model for previous categories by selecting key samples, while the latter is used to allow the model to adapt quickly to new categories. Meanwhile, a weighted loss function based on cross-entropy and distillation loss is presented in order to fine-tune the model. We demonstrate the effectiveness of the proposed CES-MOLPC method through extensive experiments on the public RML2016.04c data set and the large-scale real-world ACARS signal data set. The results of the comparative experiments demonstrate that our method can achieve significant improvements over state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Highlights: [•] We propose a fast memetic algorithm to uncover community structure in networks. [•] The proposed algorithm is based on novel multi-level learning strategies at nodes, communities and network partitions levels. [•] Our algorithm need not know the number of clusters in advance. [•] Our algorithm has superior performances in speed, accuracy and stability. [ABSTRACT FROM AUTHOR]

Abstract: Multiple Kernel Learning (MKL) aims to seek a better result than single kernel learning by combining a compact set of sub-kernels. However, MKL with L1-norm easily discards the sub-kernels with complementary information and MKL with often gets the redundant solution. To address these problems, a Selective Multiple Kernel Learning (SMKL) method, inspired by Ensemble Learning (EL), is proposed. Comparing MKL with , SMKL obtains a sparse solution by a pre-selection procedure. Comparing MKL with L1-norm, SMKL preserves the sub-kernels with complementary information by guaranteeing the high discrimination and large diversity of pre-selected sub-kernels. For quantifying the discrimination and diversity of sub-kernels, a new kernel evaluation is designed. SMKL reduces the scale of MKL optimization and saves the memory storing of the sub-kernels, which extends the scale of problem that MKL could solve. Specially, a fast SMKL method using constraint is focused, which needs no MKL optimization process. It means that the memory is hardly a limitation for MKL with the large scale problem. Experiments state that our method is effective for classification. [Copyright &y& Elsevier]

Abstract: By inspiration of the granular evolutionary algorithm, a Granular Agent Evolutionary Classification (GAEC) algorithm for the classification task in data mining is proposed. The method uses the granular agent to denote the set of some examples that have similar attributions and the knowledge base guides the evolution of granular agent. Also some granular evolutionary operators are designed for classification problem. Assimilation operator, exchange operator, and differentiation operator reflect the competitive, cooperative and self-learning ability of agent, respectively. Finally, some classification rules are extracted from granular agents by some strategy to forecast the sort of new data. Empirical study contains UCI data sets, KDDCUP99 data sets and remote image recognition. The test results show that the algorithm has a good classification prediction, and only need a small price for the training time. In most UCI data sets, the performance of GAEC is better than G-NET, OCEC and C4.5, which have good performance. At the same time, some Gaussian White Noise attributes are added to these UCI data sets and the results show GACE has some anti-noise abilities. To test the scalability of GAEC, two functions along two dimensions, the number of training examples and the number of attributes are used. Also, GAEC are applied to some real world fields, intrusion detection system and remote sensing image recognition. The experiments for KDDCUP99 verify GAEC has capability to deal with massive data in real world and good predicting capability for unknown type data. At last, the accuracy rate of GAEC is also good for the remote sensing image recognition. [Copyright &y& Elsevier]

Abstract: To approximate the multivariate functions with spatial inhomogeneity, in this paper we proposed an ortho-ridgelet neural network (ORNN) model. By taking orthonormal ridgelet, which is a “true” ridgelet function different with the “classic” ridgelet, as the activation function of the hidden neurons, the network is characterized of more efficient representation of a set of functions with linear and curvilinear singularities. Although the ortho-ridgelet is initially more complicated to be understood than ridgelet, it can be easily applied to function approximation theory in a more effective manner. Moreover, we extended the ORNN model to the time-varied situation, and the time-varied ortho-ridgelet neural network (TV-ORRN) is established for time-varied function approximation, with an incremental learning of the ridgelets output layer. The theoretical analysis is presented and some experiments are taken, and the experimental results prove its efficiency in approximation of multivariate functions with linear and curvilinear singularities, in both the stationary and time-varied environment. [ABSTRACT FROM AUTHOR]

Recently the study on the theory of wavelets shows that the wavelets have not only the multi-resolution property both in frequency and time domain, but also the good approximation ability. SVMs based on the statistical learning theory are a kind of general and effective learning machines, and have described for us the nice application blueprint in machine learning domain. There exists a bottleneck problem, or the pre-selection of kernel parameter for SVMs. In this paper, the orthogonal projection kernels of father wavelet (OPFW kernels) are introduced into SVMs. In doing so SVMs based on the OPFW kernels can have good performance in both approximation and generalisation. Simultaneously the parameter pre-selection of wavelet kernels can be implemented by discrete wavelet transform. Experiments on regression estimation illustrate the approximation and generalisation ability of our method. [ABSTRACT FROM AUTHOR]

Ship detection is a challenging task for synthetic aperture radar (SAR) images. Ships have arbitrary directionality and multiple scales in SAR images. Furthermore, there is a lot of clutter near the ships. Traditional detection algorithms are not robust to these situations and easily cause redundancy in the detection area. With the continuous improvement in resolution, the traditional algorithms cannot achieve high-precision ship detection in SAR images. An increasing number of deep learning algorithms have been applied to SAR ship detection. In this study, a new ship detection network, known as the instance segmentation assisted ship detection network (ISASDNet), is presented. ISASDNet is a two-stage detection network with two branches. A branch is called an object branch and can extract object-level information to obtain positioning bounding boxes and classification results. Another branch called the pixel branch can be utilized for instance segmentation. In the pixel branch, the designed global relational inference layer maps the features to interaction space to learn the relationship between ship and background. The global reasoning module (GRM) based on global relational inference layers can better extract the instance segmentation results of ships. A mask assisted ship detection module (MASDM) is behind the two branches. The MASDM can improve detection results by interacting with the outputs of the two branches. In addition, a strategy is designed to extract the mask of SAR ships, which enables ISASDNet to perform object detection training and instance segmentation training at the same time. Experiments carried out two different datasets demonstrated the superiority of ISASDNet over other networks. [ABSTRACT FROM AUTHOR]

Recently, with the popularity of space-borne earth satellites, the resolution of high-resolution panchromatic (PAN) and multispectral (MS) remote sensing images is also increasing year by year, multiresolution remote sensing classification has become a research hotspot. In this paper, from the perspective of deep learning, we design a dual-branch interactive spatial-channel collaborative attention enhancement network (SCCA-net) for multiresolution classification. It aims to combine sample enhancement and feature enhancement to improve classification accuracy. In the part of sample enhancement, we propose an adaptive neighbourhood transfer sampling strategy (ANTSS). Different from the traditional pixel-centric sampling strategy with orthogonal sampling angle, our algorithm allows each patch to adaptively transfer the neighbourhood range by finding the homogeneous region of the pixel to be classified. And it also adaptively adjust the sampling angle according to the texture distribution of the homogeneous region to capture neighbourhood information that is more conducive for classification. Moreover, in the part of feature enhancement part, we design a local spatial attention module (LSA-module) for PAN data to highlight the spatial resolution advantages and a global channel attention module (GCA-module) for MS data to improve the multi-channel representation. It not only highlights the spatial resolution advantage of PAN data and the multi-channel advantage of MS data, but also improves the difference between features through the interaction between the two modules. Quantitative and qualitative experimental results verify the robustness and effectiveness of the method. [ABSTRACT FROM AUTHOR]