Search

Recently, many state-of-the-art results for remote sensing image scene classification have been achieved by convolutional neural networks (CNNs) due to their large learning capability. However, in the forward process of CNNs, the high-frequency/texture features are gradually blurred with hierarchical downsampling and convolution operations. High-frequency features are important to capture the diversity within a class and the similarity between classes. For example, the line features are crucial to distinguish a tennis court from a basketball court. For tennis court in different scenes, the highlight of line features can effectively avoid the influence of diverse backgrounds. As a consequence, we propose a Laplacian high-frequency convolutional block (LHCB) based on CNN to extract useful high-frequency features by trainable Laplacian operator. To propagate high-frequency features, we embed LHCB into the existing CNN structures and obtain LHNet. In LHNet, there are two pathways. The original CNN architecture can be taken as the low-frequency pathway, and we propose a high-frequency pathway (HFP) based on LHCB that propagates the residual high-frequency features blurred in each low-frequency layer. Considering that the high-frequency features usually show large variance between images of the same class, we propose a new objective for HFP to enhance the intraclass similarity of high-frequency features. The final objective function is obtained by combining the new objective and the baseline classification objective. Numerous experiments on three public available remote sensing image scene classification datasets, the Northwestern Polytechnical University-Remote Sensing Image Scene Classification (NWPU-RESISC45), Aerial Image Dataset (AID), and the University of California at Merced (UC Merced), demonstrate the superior performance of the proposed method. [ABSTRACT FROM AUTHOR]

Remote sensing object tracking (RSOT) is a novel and challenging problem due to the negative effects of weak features and background noise. In this article, from the perspective of attention-focus deep learning, we propose a joint Siamese attention-aware network (JSANet) for efficient remote sensing tracking which contains both the self-attention and cross-attention modules. First, the self-attention modules we propose emphasize the interdependent channel-wise coefficient via channel attention and conduct corresponding space transformation of spatial domain information with spatial attention. Second, the cross-attention is designed to aggregate rich contextual interdependencies between the Siamese branches via channel attention and excavate association produces reliable correspondence with spatial attention. In addition, a composite feature combine strategy is designed to fuse multiple attention features. Experimental results on the Jilin-1 satellite video datasets demonstrate that the proposed JSANet achieves state-of-the-art performance in terms of precision and success rate, demonstrating the effectiveness of the proposed methods. [ABSTRACT FROM AUTHOR]

Semantic segmentation is an important yet unsolved problem in aerial scenes understanding. One of the major challenges is the intense variations of scenes and object scales. In this article, we propose a novel multi-scale aware-relation network (MANet) to tackle this problem in remote sensing. Inspired by the process of human perception of multi-scale (MS) information, we explore discriminative and diverse MS representations. For discriminative MS representations, we propose an inter-class and intra-class region refinement (IIRR) method to reduce feature redundancy caused by fusion. IIRR utilizes the refinement maps with intra-class and inter-class scale variation to guide MS fine-grained features. Then, we propose multi-scale collaborative learning (MCL) to enhance the diversity of MS feature representations. The MCL constrains the diversity of MS feature network parameters to obtain diverse information. Also, the segmentation results are rectified according to the dispersion of the multilevel network predictions. In this way, MANet can learn MS features by collaboratively exploiting the correlation among different scales. Extensive experiments on image and video datasets, which have large-scale variations, have demonstrated the effectiveness of our proposed MANet. [ABSTRACT FROM AUTHOR]

Supervised change detection methods always face a big challenge that the current scene (target domain) is fully unlabeled. In remote sensing, it is common that we have sufficient labels in another scene (source domain) with a different but related data distribution. In this article, we try to detect changes in the target domain with the help of the prior knowledge learned from multiple source domains. To achieve this goal, we propose a change detection framework based on selective adversarial adaptation. The adaptation between multisource and target domains is fulfilled by two domain discriminators. First, the first domain discriminator regards each scene as an individual domain and is designed for identifying the domain to which each input sample belongs. According to the output of the first domain discriminator, a subset of important samples is selected from multisource domains to train a deep neural network (DNN)-based change detection model. As a result, not only the positive transfer is enhanced but also the negative transfer is alleviated. Second, as for the second domain discriminator, all the selected samples are thought from one domain. Adversarial learning is introduced to align the distributions of the selected source samples and the target ones. Consequently, it further adapts the knowledge of change from the source domain to the target one. At the fine-tuning stage, target samples with reliable labels and the selected source ones are used to jointly fine-tune the change detection model. As the target domain is fully unlabeled, homogeneity- and boundary-based strategies are exploited to make the pseudolabels from a preclassification map reliable. The proposed method is evaluated on three SAR and two optical data sets, and the experimental results have demonstrated its effectiveness and superiority. [ABSTRACT FROM AUTHOR]

Supervised deep neural networks (DNNs) have been extensively used in diverse tasks. Generally, training such DNNs with superior performance requires a large amount of labeled data. However, it is time-consuming and expensive to manually label the data, especially for tasks in remote sensing, e.g., change detection. The situation motivates us to resort to the existing related images with labels, from which the concept of change can be adapted to new images. However, the distributions of the related labeled images (source domain) and unlabeled new images (target domain) are similar but not identical. It impedes a change detection model learned from source domains being well applied to the target domain. In this paper, we propose a transferred deep learning-based change detection framework to solve this problem. It consists of pretraining and fine-tuning stages. In the pretraining process, we propose two tasks to be learned simultaneously, namely, change detection for the source domain with labels and reconstruction of the unlabeled target data. The auxiliary task aims to reconstruct the difference image (DI) for the target domain. DI is an effective feature, such that the auxiliary task is of much relevance to change detection. The lower layers are shared between these two tasks in the training process. It mitigates the distribution discrepancy between the source and target domains and makes the concept of change from the source domain adapt to the target domain. In addition, we evaluate three modes of the U-net architecture to merge the information for a pair of patches. To fine-tune the change detection network (CDN) for the target domain, two strategies are exploited to select the pixels that have a high possibility of being correctly classified by an unsupervised approach. The proposed method demonstrates an excellent capacity for adapting the concept of change from the source domain to the target domain. It outperforms the state-of-the-art change detection methods via experimental results on real remote sensing data sets. [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]

Image reranking is a popular postprocessing method for remote sensing image retrieval (RSIR), which aims at enhancing the initial retrieval performance. In general, it takes either users’ opinions or the relationships between images into consideration to find an optimal reranked list based on the initial retrieved results. In this paper, we present a reranking method for improving RSIR, which is named two-stage reranking (TSR). Suppose the k -nearest neighbors of a query RS image have been obtained by the initial retrieval. The first step of our TSR is to edit these neighbors using the editing scheme. A handful of informative and representative RS images are selected by the active learning algorithm, and their binary labels are provided by the users relative to the query image. Then, a binary classifier is trained using the selected RS images and their labels to classify the rest of the neighbors. Finally, both classification results and rank information in the initial retrieval results are considered to decide which neighbor should be excluded. In the next step, the remaining RS images are reranked by the proposed reranking scheme, i.e., multisimilarity fusion reranking. Both the user’s experience and image relationships are taken into account in TSR to ensure the performance of the reranking. The efficiency and the robustness of our method are validated by experiments conducted on two different types of RS images. Compared with the existing visual reranking approaches, our method achieves improved performance. [ABSTRACT FROM PUBLISHER]

An approach of weighted Wishart distance learning, shorted for W2-based distance learning, is proposed for polarimetric synthetic aperture radar (PolSAR) image classification. It aims to adjust the Wishart distance by enhancing discrimination as well as exploiting spatial information. The proposed distance learning keeps samples within the same category close and separates samples from the different classes far apart. It is effectively implemented by solving a linear programming. Input of W2-based distance learning is called weighted Wishart feature, which is designed specifically for PolSAR data to describe the Wishart distribution, achieve regional consistency, and reduce speckle noise. Weight is calculated according to an adaptive window, where homogeneous samples are derived based on a connected region and extracted edge information. With this feature, W2-based distance learning is a whole scheme to adjust the Wishart distance. Furthermore, our experiments with benchmark data sets suggest that the proposed scheme provides both improved performance in terms of visual effect and classification accuracy. The achieved overall accuracy is better by more than 7% compared to other state-of-art methods. [ABSTRACT FROM PUBLISHER]

The multispectral (MS) and the panchromatic (PAN) images represent complementary and synergistic spatial spectral information, how to make optimal use of the advantages of them has become a hot research topic. This paper proposes a selectable Transformer and Gist CNN network (STGC-Net). It designs a subspace similar recombination module (SSR-Module) based on non-negative matrix factorization (NMF) and the self-attention mechanism for feature decomposition. This can alleviate the redundant information of multi-modal data and extract their own singular and common features. Considering that the MS and the PAN images exhibit different advantageous properties, a selectable self-attention spectral feature extraction module (S 3 FE-Module) and a multi-stream Gist spatial feature extraction module (MGSFE-Module) are proposed for the different singular features. The former can refine the Transformer's input and simultaneously characterize the sequence information between channels for the MS image. The latter introduces the positional relationship between local features while extracting spatial features for the PAN image, thereby improving the accuracy of scene classification. Experimental results indicate that the proposed method performs better than the other methods. The relevant code of this paper is provided at: https://github.com/ru-willow/ST-GC-Net. • A subspace similar recombination module is proposed. • The selectable self-attention spectral feature extraction module is designed. • The multi-stream Gist spatial feature extraction module is introduced. [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]

Sparse representation-based classification (SRC) and collaborative representation-based classification (CRC) have been proposed for image classification. Sparsity is overemphasized in SRC, but ignored in CRC, which has low representation complexity. In this article, a new local collaborative representation scheme is proposed for the classification of hyperspectral imagery (HSI). First, for a test sample with an unknown label, a few significant atoms are selected based on their contributions in representing the test sample, where the-norm is replaced by the-norm like CRC to reduce the complexity. Next, the test sample is represented by the active atoms. Finally, the class label of the test sample is determined based on the contribution of each class of active atoms. Additionally, in order to improve the robustness of the active atoms, both the test sample and the samples in its neighbourhood as context information are considered in the selection of active atoms and the class determination. Experimental results on three real HSI data sets confirm the effectiveness and accuracy of the proposed method. [ABSTRACT FROM AUTHOR]

Deep convolutional neural networks (DCNNs) are driving progress in object detection of high-resolution remote sensing images. Region proposal generation, as one of the key steps in object detection, has also become the focus of research. High-resolution remote sensing images usually contain various sizes of objects and complex background, small objects are easy to miss or be mis-identified in object detection. If the recall rate of region proposal of small objects and multi-scale objects can be improved, it will bring an improvement on the performance of the accuracy in object detection. Spatial attention is the ability to focus on local features in images and can improve the learning efficiency of DCNNs. This study proposes a multi-scale spatial attention region proposal network (MSA-RPN) for high-resolution optical remote sensing imagery. The MSA-RPN is an end-to-end deep learning network with a backbone network of ResNet. It deploys three novel modules to fulfill its task. First, the Scale-specific Feature Gate (SFG) focuses on features of objects by processing multi-scale features extracted from the backbone network. Second, the spatial attention-guided model (SAGM) obtains spatial information of objects from the multi-scale attention maps. Third, the Selective Strong Attention Maps Model (SSAMM) adaptively selects sliding windows according to the loss values from the system's feedback, and sends the windowed samples to the spatial attention decoder. Finally, the candidate regions and their corresponding confidences can be obtained. We evaluate the proposed network in a public dataset LEVIR and compare with several state-of-the-art methods. The proposed MSA-RPN yields a higher recall rate of region proposal generation, especially for small targets in remote sensing images. [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]

Water area segmentation in remote sensing is of great importance for flood monitoring. To overcome some challenges in this task, we construct the Water Index and Polarization Information (WIPI) multi-modality dataset and propose a multi-Modality Fusion and Gated multi-Filter U-Net (MFGF-UNet) convolutional neural network. The WIPI dataset can enhance the water information while reducing the data dimensionality: specifically, the Cloud-Free Label provided in the dataset can effectively alleviate the problem of labeled sample scarcity. Since a single form or uniform kernel size cannot handle the variety of sizes and shapes of water bodies, we propose the Gated Multi-Filter Inception (GMF-Inception) module in our MFGF-UNet. Moreover, we utilize an attention mechanism by introducing a Gated Channel Transform (GCT) skip connection and integrating GCT into GMF-Inception to further improve model performance. Extensive experiments on three benchmarks, including the WIPI, Chengdu and GF2020 datasets, demonstrate that our method achieves favorable performance with lower complexity and better robustness against six competing approaches. For example, on the WIPI, Chengdu and GF2020 datasets, the proposed MFGF-UNet model achieves F1 scores of 0.9191, 0.7410 and 0.8421, respectively, with the average F1 score on the three datasets 0.0045 higher than that of the U-Net model; likewise, GFLOPS were reduced by 62% on average. The new WIPI dataset, the code and the trained models have been released on GitHub. [ABSTRACT FROM AUTHOR]

In the era of big data, making full use of remote sensing images to automatically extract surface water bodies (WBs) in complex environments is extremely challenging. Due to the weak capability of existing algorithms in extracting small WBs and WB edge information from remote sensing images, we proposed a new method—Multiscale Fusion SegFormer (MF-SegFormer)—for WB extraction in the Weihe River Basin of China using Landsat 8 OLI images. The MF-SegFormer method adopts a cascading approach to fuse features output by the SegFormer encoder at multiple scales. A feature fusion (FF) module is proposed to enhance the extraction of WB edge information, while an Atrous Spatial Pyramid Pooling (ASPP) module is employed to enhance the extraction of small WBs. Furthermore, we analyzed the impact of four kinds of band combinations on WB extraction by the MF-SegFormer model, including true color composite images, false color images, true color images, and false color images enhanced by Gaussian stretch. We also compared our proposed method with several different approaches. The results suggested that false color composite images enhanced by Gaussian stretching are beneficial for extracting WBs, and the MF-SegFormer model achieves the highest accuracy across the study area with a precision of 77.6%, recall of 84.4%, F1-score of 80.9%, and mean intersection over union (mIoU) of 83.9%. In addition, we used the determination coefficient (R2) and root-mean-square error (RMSE) to evaluate the performance of river width extraction. Our extraction results in an overall R2 of 0.946 and an RMSE of 28.21 m for the mainstream width in the "Xi'an-Xianyang" section of the Weihe River. The proposed MF-SegFormer method used in this study outperformed other methods and was found to be more robust for WB extraction. [ABSTRACT FROM AUTHOR]

Remote sensing image scene classification (RSISC) has garnered significant attention in recent years. Numerous methods have been put forward in an attempt to tackle this issue, particularly leveraging deep learning methods that have shown promising performance in classifying remote sensing image (RSI). However, it is widely recognized that deep learning methods typically require a substantial amount of labeled data to effectively converge. Acquiring a sufficient quantity of labeled data often necessitates significant human and material resources. Hence, few-shot RSISC has become highly meaningful. Fortunately, the recently proposed deep nearest neighbor neural network based on the attention mechanism (DN4AM) model incorporates episodic training and class-related attention mechanisms, effectively reducing the impact of background noise regions on classification results. Nevertheless, the DN4AM model does not address the problem of significant intra-class variability and substantial inter-class similarities observed in RSI scenes. Therefore, the discriminative enhanced attention-based deep nearest neighbor neural network (DEADN4) is proposed to address the few-shot RSISC task. Our method makes three contributions. Firstly, we introduce center loss to enhance the intra-class feature compactness. Secondly, we utilize the deep local-global descriptor (DLGD) to increase inter-class feature differentiation. Lastly, we modify the Softmax loss by incorporating cosine margin to amplify the inter-class feature dissimilarity. Experiments are conducted on three diverse RSI datasets to gauge the efficacy of our approach. Through comparative analysis with various cutting-edge methods including MatchingNet, RelationNet, MAML, Meta-SGD, DN4, and DN4AM, our approach showcases promising outcomes in the few-shot RSISC task. [ABSTRACT FROM AUTHOR]

With the development of remote sensing earth observation technology, object tracking has gained attention for its broad application prospects in computer vision. However, object tracking is challenging owing to the background clutter, occlusion, and scale variation that often appear in remote sensing videos. Many existing trackers cannot accurately track the object for remote sensing videos with complex backgrounds. Several tracking methods can handle just one situation, such as occlusion. In this article, we propose a Siamese multi-scale adaptive search (SiamMAS) network framework to achieve object tracking for remote sensing videos. First, a multi-scale cross correlation is presented to obtain a more discriminative model and comprehensive feature representation, improving the performance of the model to handle complex backgrounds in remote sensing videos. Second, an adaptive search module is employed that augments the Kalman filter with a partition search strategy for object motion estimation. The Kalman filter is adopted to re-detect the object when the network cannot track the object in the current frame. Moreover, the partition search strategy can help the Kalman filter accomplish a more accurate region-proposal selection. Finally, extensive experiments on remote sensing videos taken from Jilin-1 commercial remote sensing satellites show that the proposed tracking algorithm achieves strong tracking performance with 0.913 precision while running at 37.528 frames per second (FPS), demonstrating its effectiveness and efficiency. [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]

Recently, a massive number of deep learning-based approaches have been successfully applied to various remote sensing image (RSI) recognition tasks. However, most existing advances of deep learning methods in the RSI field heavily rely on the features extracted by the manually designed backbone network, which severely hinders the potential of deep learning models due to the complexity of RSI and the limitation of prior knowledge. In this paper, we research a new design paradigm for the backbone architecture in RSI recognition tasks, including scene classification, land-cover classification and object detection. A novel one-shot architecture search framework based on a weight-sharing strategy and an evolutionary algorithm is proposed, called RSBNet, which consists of three stages: Firstly, a supernet constructed in a layer-wise search space is pretrained on a self-assembled large-scale RSI dataset based on an ensemble single-path training strategy. Next, the pre-trained supernet is equipped with different recognition heads through the switchable recognition module and respectively fine-tuned on the target dataset to obtain task-specific supernet. Finally, we search for the optimal backbone architecture for different recognition tasks based on the evolutionary algorithm without any network training. Extensive experiments have been conducted on five benchmark datasets for different recognition tasks, the results show the effectiveness of the proposed search paradigm and demonstrate that the searched backbone is able to flexibly adapt different RSI recognition tasks and achieve impressive performance. [ABSTRACT FROM AUTHOR]

Sparse representation has been widely used in the field of remote sensing image super-resolution (SR) to restore a high-quality image from a low-resolution (LR) image, e.g., from the blurred and downsampled version of an LR image’s high-resolution (HR) counterpart. It is well known that each image patch can be represented by a linear combination of the atoms of an overcomplete dictionary, and we can obtain an expression of sparse coefficients by $l_{1}$ norm regularization. Owing to the lack of an inner relationship between image patches and an image’s global information, the traditional methods of jointly training two overcomplete dictionaries cannot obtain good SR results. Therefore, we propose an effective approach for remote sensing image SR based on sparse representation. More specifically, a novel global joint dictionary model (GJDM) is used to explore the prior knowledge of images, including local and global characteristics. First, we train two dictionaries for detail image patches and HR patches. Second, in order to enhance the inner relationship between image patches, we introduce a global self-compatibility model for global regularization. Finally, the sparse representation and the local and nonlocal constraints are integrated to improve the performance of the model, and the fast adaptive shrinkage-thresholding algorithm is employed to solve the convex optimization problem in the GJDM. Compared with other methods, the results of the proposed method show good SR performance in preserving details and texture information and significant improvement in a peak signal-to-noise ratio. [ABSTRACT FROM AUTHOR]

Detecting changed regions between two given synthetic aperture radar images is very important to monitor change of landscapes, change of ecosystem and so on. This can be formulated as a classification problem and addressed by learning a classifier, traditional machine learning classification methods very easily stick to local optima which can be caused by noises of data. Hence, we propose an unsupervised algorithm aiming at constructing a classifier based on self-paced learning. Self-paced learning is a recently developed supervised learning approach and has been proven to be capable to overcome effectively this shortcoming. After applying a pre-classification to the difference image, we uniformly select samples using the initial result. Then, self-paced learning is utilized to train a classifier. Finally, a filter is used based on spatial contextual information to further smooth the classification result. In order to demonstrate the efficiency of the proposed algorithm, we apply our proposed algorithm on five real synthetic aperture radar images datasets. The results obtained by our algorithm are compared with five other state-of–the-art algorithms, which demonstrates that our algorithm outperforms those state-of-the-art algorithms in terms of accuracy and robustness. [ABSTRACT FROM AUTHOR]

Remote sensing image scene classification has become more and more popular in recent years. As we all know, it is very difficult and time-consuming to obtain a large number of manually labeled remote sensing images. Therefore, few-shot scene classification of remote sensing images has become an urgent and important research task. Fortunately, the recently proposed deep nearest neighbor neural network (DN4) has made a breakthrough in few-shot classification. However, due to the complex background in remote sensing images, DN4 is easily affected by irrelevant local features, so DN4 cannot be directly applied in remote sensing images. For this reason, a deep nearest neighbor neural network based on attention mechanism (DN4AM) is proposed to solve the few-shot scene classification task of remote sensing images in this paper. Scene class-related attention maps are used in our method to reduce interference from scene-semantic irrelevant objects to improve the classification accuracy. Three remote sensing image datasets are used to verify the performance of our method. Compared with several state-of-the-art methods, including MatchingNet, RelationNet, MAML, Meta-SGD and DN4, our method achieves promising results in the few-shot scene classification of remote sensing images. [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]

Recently few-shot object detection (FSOD) in remote sensing images (RSIs) has drawn increasing attention. However, the current FSOD methods in RSIs merely focus on the detection performance of few-shot novel classes while ignoring the severe degradation of the base class performance. Generalized few-shot object detection (G-FSOD) aims to solve the FSOD problem without forgetting previous knowledge. In this paper, we focus on the G-FSOD in RSIs and propose a Generalized Few-Shot Detector (G-FSDet) that can learn novel knowledge without forgetting. Through the comprehensive analysis of each component in the detector, a novel efficient transfer-learning framework is presented as the foundation of our G-FSDet, which is more suitable for FSOD in remote sensing scenes. Considering the greater intra-class diversity and lower inter-class separability of geospatial objects, we design a metric-based discriminative loss to learn a more discriminative classifier in the few-shot fine-tuning stage. Furthermore, a representation compensation module is proposed to alleviate the catastrophic forgetting problem by decoupling the representation learning of previous and novel knowledge. Extensive experiments on DIOR and NWPU VHR-10.v2 datasets demonstrate that our proposed G-FSDet achieves competitive novel class performance with minor degradation in the base class, reaching state-of-the-art overall performance among all few-shot settings. The source code is available at (https://github.com/RSer-XDU/G-FSDet). [ABSTRACT FROM AUTHOR]

Deep descriptors have advantages over handcrafted descriptors on local image patch matching. However, due to the complex imaging mechanism of remote sensing images and the significant differences in appearance between multi-modal images, existing deep learning descriptors are unsuitable for multi-modal remote sensing image registration directly. To solve this problem, this article proposes a deep feature correlation learning network (Cnet) for multi-modal remote sensing image registration. First, Cnet builds a feature learning network based on the deep convolutional network with the attention learning module, to enhance feature representation by focusing on meaningful features. Second, this article designs a novel feature correlation loss function for Cnet optimization. It focuses on the relative feature correlation between matching and nonmatching samples, which can improve the stability of network training and decrease the risk of overfitting. In addition, the proposed feature correlation loss with a scale factor can further enhance network training and accelerate network convergence. Extensive experimental results on image patch matching (Brown, HPatches), cross-spectral image registration (VIS-NIR), multi-modal remote sensing image registration, and single-modal remote sensing image registration have demonstrated the effectiveness and robustness of the proposed method. [ABSTRACT FROM AUTHOR]

In recent years, convolutional neural network (CNN)-based methods have been widely used for remote sensing (RS) scene classification tasks and have achieved excellent results. However, CNNs are not good at exploring contextual information, which is essential for fully understanding RS scenes. A new model named transformer attracts researchers’ attention to address this problem, which is skilled in mining the latent contextual information in RS scenes. Nevertheless, since the contents of RS scenes are diverse in type and various in scale, the performance of the original transformer in RS scene classification cannot reach what we expect. In addition, due to the specific self-attention mechanism, the time costs of the transformer are high, which hinders its practicability in the RS community. To overcome the above limitations, we propose a new model named efficient multiscale transformer and cross-level attention learning (EMTCAL) for RS scene classification in this article. EMTCAL combines the advantages of CNN and transformer to mine information within RS scenes fully. First, it uses a multilayer feature extraction module (MFEM) to acquire global visual features and multilevel convolutional features from RS scenes. Second, a contextual information extraction module (CIEM) is proposed to capture rich contextual information from multilevel features. In CIEM, taking the characteristics of RS scenes and the computational complexity into account, we propose an efficient multiscale transformer (EMST). EMST can mine the abundant knowledge with various scales hidden in RS scenes and model their inherent relations at small time costs. Third, a cross-level attention module (CLAM) is developed to aggregate and explore correlations of multilevel features. Finally, a class score fusion module (CSFM) is designed to integrate the contributions of global and aggregated multilevel features for discriminative scene representations. Extensive experiments are conducted on three public RS scene datasets. The positive results demonstrate that our EMTCAL can achieve superior classification performance and outperform many state-of-the-art methods. Our source codes are available in https://github.com/TangXu-Group/Remote-Sensing-Images-Classification/tree/main/EMTCAL. [ABSTRACT FROM AUTHOR]

During the past decades, convolutional neural network (CNN)-based models have achieved notable success in remote sensing image classification due to their powerful feature representation ability. However, the lack of explainability during the decision-making process is a common criticism of these high-capacity networks. Local explanation methods that provide visual saliency maps have attracted increasing attention as a means to surmount the barrier of explainability. However, the vast majority of research is conducted on the last convolutional layer, where the salient regions are unintelligible for partial remote sensing images, especially scenes that contain plentiful small targets or are similar to the texture image. To address these issues, we propose a novel framework called Prob-POS, which consists of the class-activation map based on the probe network (Prob-CAM) and the weighted probability of occlusion (wPO) selection strategy. The proposed probe network is a simple but effective architecture to generate elaborate explanation maps and can be applied to any layer of CNNs. The wPO is a quantified metric to evaluate the explanation effectiveness of each layer for different categories to automatically pick out the optimal explanation layer. Variational weights are taken into account to highlight the high-scoring regions in the explanation map. Experimental results on two publicly available datasets and three prevalent networks demonstrate that Prob-POS improves the faithfulness and explainability of CNNs on remote sensing images. [ABSTRACT FROM AUTHOR]

Optical and synthetic aperture radar (SAR) image registration is important for multimodal remote sensing image information fusion. Recently, deep matching networks have shown better performances than traditional methods of image matching. However, due to significant differences between optical and SAR images, the performances of existing deep learning methods still need to be further improved. This article proposes a self-distillation feature learning network (SDNet) for optical and SAR image registration, improving performance from network structure and network optimization. First, we explore the impact of different weight-sharing strategies on optical and SAR image matching. Then, we design a partially unshared feature learning network for multimodal image feature learning. It has fewer parameters than the fully unshared network and has more flexibility than the fully shared network. In addition, the limited binary supervised information (matching or nonmatching) is insufficient to train the deep matching networks for optical-SAR image registration. Thus, we propose a self-distillation feature learning method to exploit more similarity information for deep network optimization enhancement, such as the similarity ordering between a series of nonmatching patch pairs. The exploited rich similarity information will significantly enhance network training and improve matching accuracy. Finally, existing deep learning methods brute-force make the matching features of the optical and SAR image patches similar, which will lead to the loss of discriminative information and degeneration of the matching performances. Thus, we build an auxiliary task reconstruction learning to optimize the feature learning network to keep more discriminative information. Extensive experiments demonstrate the effectiveness of our proposed method on multimodal image registration. [ABSTRACT FROM AUTHOR]

Change detection, as an important task of remote sensing image processing, has a wide range of applications in many aspects such as land use and natural disaster assessment. Recent change detection methods have achieved good results. However, due to the environmental difference between the bi-temporal images and the complicated imaging condition, there are usually problems such as missing small objects, incomplete objects, and rough edges in the change detection results. The existing change detection methods usually lack attention in these areas. In this paper, we propose a Siamese change detection method, named SMD-Net, for bi-temporal remote sensing change detection. The proposed model uses multi-scale difference maps to enhances the information of the changed areas step by step in order to have better change detection results. Furthermore, we propose a Siamese residual multi-kernel pooling module (SRMP) for high-level features to enhance the high-level change information of the model. For the low-level features of multiple skip connections, we propose a feature difference module (FDM) that uses feature difference to fully extract the change information and help the model generate more accurate details. The experimental results of our method on three public datasets show that compared with other benchmark methods, our network comprises better effectiveness and has a better trade-off between accuracy and calculation cost. [ABSTRACT FROM AUTHOR]

Content-based remote sensing (RS) image retrieval (CBRSIR) is a critical way to organize high-resolution RS (HRRS) images in the current big data era. The increasing volume of HRRS images from different satellites and sensors leads to more attention to the cross-source CSRSIR (CS-CBRSIR) problem. Due to the data drift, one crucial problem in CS-CBRSIR is the modality discrepancy. Most existing methods focus on finding a common feature space for various HRRS images to address this issue. In this space, their similarity relations can be measured directly to obtain the cross-source retrieval results straight. This way is feasible and reasonable, however, the specific information corresponding to HRRS images from different sources is always ignored, limiting retrieval performance. To overcome this limitation, we develop a new model for CS-CBRSIR in this paper named dual modality collaborative learning (DMCL). To fully explore the specific information from diverse HRRS images, DMCL first introduces ResNet50 as the feature extractor. Then, a common space mutual learning module is developed to map the specific features into a common space. Here, the modality discrepancy is reduced from the aspects of features and their distributions. Finally, to supplement the specific knowledge to the common features, we develop modality transformation and the dual-modality feature learning modules. Their function is to transmit the specific knowledge from different sources mutually and fuse the specific and common features adaptively. The comprehensive experiments are conducted on a public dataset. Compared with many existing methods, the behavior of our DMCL is stronger. These encouraging results for a public dataset indicate that the proposed DMCL is useful in CS-CBRSIR tasks. [ABSTRACT FROM AUTHOR]

Recently, multicategory object detection in high-resolution remote sensing images is still a challenge. First, objects with significant scale differences exist in one scene simultaneously, so it is generally difficult for the detectors to balance the detection performance of large and small objects. Second, because of the complex background and the objects’ densely distributed characteristics in the remote sensing images, the extracted features usually have noise and blurred boundaries, which interfere with the detection performance of the object detectors. With this observation, we propose an end-to-end scale-aware network called feature split–merge–enhancement network (SME-Net) for remote sensing object detection, composed of the feature split-and-merge (FSM) module, the offset-error rectification (OER) module, and the object saliency enhancement (OSE) strategy. FSM eliminates salient information of large objects to highlight the features of small objects in the shallow feature maps. It also transmits the effective detailed features of large objects to the deep feature maps, alleviating feature confusion between multiscale objects. OER corrects the inconsistency of the features spatial layout among the multilayer feature maps by the proposed offset loss, so as to achieve supervised elimination and transmission in FSM. OSE enhances the features of interests and suppresses the background information by the proposed membership function, thus preventing false detection and missed detection caused by noise and blurred boundaries. The effectiveness of the proposed algorithm has been verified on multiple datasets. Our code is available at: https://github.com/Momuli/SMENet.git [ABSTRACT FROM AUTHOR]

With the explosive growth of the volume and resolution of high-resolution remote-sensing (HRRS) images, the management of them becomes a challenging task. The traditional content-based remote-sensing image retrieval (CBRSIR) technologies cannot meet what we expect due to the large volume of image archives and complex contents within HRRS images. As a successful approximate nearest neighborhood (ANN) search technique, Hash learning has received wide attention, especially when deep convolutional neural networks (DCNNs) appear. Due to DCNNs’ strong capacity of feature learning, many DCNN-based hashing methods have been proposed and achieved good performance for large-scale CBRSIR tasks. Nevertheless, their limitation is that a large of labeled training samples should be collected for training the deep models. To overcome this limitation, this article, therefore, develops a new supervised hash learning method for the large-scale HRRS CBRSIR task based on meta-learning, which could achieve well-retrieval performance with a few labeled training samples. First, taking the characteristics of HRRS into account, we develop a self-adaptive convolution (SAP-Conv) block and design a hashing net based on the block. SAP-Conv can learn robust features from HRRS images by exploring their multiscale information. Second, to enhance the generalization of the hashing net under a few labeled training samples, the hash learning is formulated in a meta-way, and we name it meta-hashing. Meta-hashing can effectively preserve the similarities between support and query set, and the similarities between samples within support set by the developed loss function. To further improve the performance of meta-hashing, we expand it to a dynamic version named dynamic-meta-hashing, in which the numbers of support and query are changeable in the training phase. Experimental results counted on the three widely used HRRS datasets demonstrate our dynamic-meta-hashing and meta-hashing can achieve promising performance in large-scale HRRS CBRSIR tasks based on a few training samples. Our source codes are available at https://github.com/TangXu-Group/Meta-hashing. [ABSTRACT FROM AUTHOR]

For remote sensing image (RSI) captioning tasks, two-stage RSI captioning methods have achieved high performance because they introduce the results of other RSI tasks, such as image classification as prior information. However, most previous works treat image classification in the two-stage RSI captioning method as a single-label classification task, which cannot adequately describe the entire contents of complex RSIs and may cause semantic ambiguity. To settle this problem and further refine image feature representation, we introduce multi-label classification into two-stage RSI captioning to provide sufficient and accurate prior semantic information and propose a multi-label semantic feature fusion (MLSFF) framework. Specifically, we design a robust multi-label semantic attribute extractor to extract multi-label semantic attributes of RSIs. To obtain discriminative feature representations for RSI captioning, we propose two cross-modal semantic feature fusion operators that fuse the extracted semantic attributes and the image feature extracted by the convolutional neural network. The results of extensive numerical experiments show that the proposed method can achieve state-of-the-art performance on the UCM-Captions, Sydney-Captions, and RSICD datasets. Specifically, on the UCM-Captions dataset, our method achieves a gain of 8.2% in S m score over the SAT (LAM) method (Zhang et al., 2019c). On the Sydney-Captions dataset, our method improves the S m score by 17.4% compared with the TCE loss-based method (Li et al., 2020a). On the RSICD, our method outperforms the multi-level attention method (Li et al., 2020b) by 3.2% in terms of the S m score. Code is available at https://github.com/xtye5025/MLSFF. [ABSTRACT FROM AUTHOR]

Ship detection plays a significant role in the high-resolution remote sensing (HRRS) community, but it is a challenging task due to the complex contents within HRRS images and the diverse orientation of ships. Recently, with the development of deep learning, the performance of the HRRS ship detection model has been improved greatly. Most of them employ deep networks and complicate anchor mechanism to get well ship detection results. Nevertheless, this kind of combination limits the detection efficiency. To address this problem, a new approach named accurate and real-time rotational ship detector (AR2Det) is proposed in this article to detect ships without the anchor mechanism. Based on the extracted features by the feature extraction module (FEM) and the central information of ships, AR2Det adopts two simple modules, ship detector (SDet) and center detector (CDet), to generate and improve the detection results, respectively. AR2Det is efficient due to the simple postprocessing and the lightweight network. Also, AR2Det performs satisfactorily due to the effective generation and enhancement strategy of bounding boxes. The extensive experiments are conducted on a public HRRS image ship detection dataset HRSC2016. The promising results show that our method outperforms the state-of-the-art approaches in terms of both accuracy and speed. [ABSTRACT FROM AUTHOR]

Convolutional neural network (CNN)-based research has been successfully applied in remote sensing image classification due to its powerful feature representation ability. However, these high-capacity networks bring heavy inference costs and are easily overparameterized, especially for the deep CNNs pretrained on natural image datasets. Network pruning is regarded as a prevalent approach for compressing networks, but most existing research ignores model interpretability while formulating pruning criterion. To address these issues, a network pruning method for remote sensing image classification based on interpretable CNNs is proposed. More specifically, an original interpretable CNN with a predefined pruning ratio is trained at first. The filters, namely channels in the high convolutional layer, are able to learn specific semantic meanings in proportion to the predefined pruning ratio. The filters without interpretability are supposed to be removed. As for other convolutional layers, a sensitivity function is designed to assess the risk of pruning channels for each layer, and furthermore, the pruning ratio for each layer is corrected adaptively. The pruning method based on the proposed sensitivity function is effective and requires little computational costs to search abandoned channels without damaging classification performance. To demonstrate the effectiveness, the proposed method is implemented on different scales of modern CNN models, including VGG-VD and AlexNet. The experimental results, obtained on the UC Merced dataset and NWPU-RESISC45 dataset, prove that our method significantly reduces the inference costs and improves the interpretability of networks. [ABSTRACT FROM AUTHOR]

Synthetic aperture radar (SAR) is widely used in the field of modern remote sensing due to its high resolution for a comparatively small antenna. However, there are still some difficulties in the processing of SAR images. In particular, accurate segmentation of small targets and image corners remains an important challenge, as these can easily be lost during conventional image smoothing and denoising methods. To address this, we propose an SAR image segmentation algorithm based on constrained smoothing and hierarchical label correction (CSHLC). First, a Canny algorithm is used to extract the edges of SAR images, and the Gaussian smoothing is performed on SAR images under edge constraints to achieve noise reduction so that the edges of small and big targets are well preserved. Second, a preliminary K-means clustering is conducted on the smoothing results, and then, a Markov random field (MRF) model is used on the clustering results (“original label” results), iteratively calculating a maximum likelihood set of pixel labels. Finally, through two label correction methods, pixel group counting comparison (PGCC) and gray similarity comparison (GSC), the labels of the MRF output are further checked and corrected to obtain final segmentation results. Compared with seven state-of-the-art algorithms, simulation results on both simulated SAR images and real SAR images show that the proposed CSHLC delivers higher accuracy while better retaining corners and small targets. [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]

Pan sharpening is used to fuse a low-resolution multispectral (MS) image and a high-resolution panchromatic (PAN) image to obtain a high-resolution MS image. This article proposes PSMD-Net, an end-to-end pan-sharpening method based on a multi-scale dense network. A shallow feature extraction layer (SFEL) extracts the shallow features from the original images, and these are used as an input to a global dense feature fusion (GDFF) network to learn the global features for image reconstruction. A multiscale dense block (MDB) is designed to fully extract the spatial and spectral information from the shallow features in the GDFF network. In the proposed network, multiple MDBs are stacked to extract rich, multi-scale dense hierarchical features, and a global dense connection (GDC) is designed to allow direct connections from the state of the current MDB to all subsequent MDBs to extract more advanced features. The extracted hierarchical features are sent to the global feature fusion layer (GFFL) to adaptively learn the global features for image reconstruction. Finally, global residual learning (GRL) is adopted to force the network to pay more attention to the changing part of the image. We perform experiments on simulated and real data from WorldView-2 and WorldView-3 satellites. Visual and quantitative assessment results demonstrate that PSMD-Net yields higher-resolution fusion images than the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In recent years, the panchromatic (PAN) and the multispectral (MS) remote sensing images classification has become a research hotspot. In this paper, we propose a spatial-channel progressive fusion residual network (SCPF-ResNet) for multi-resolution remote sensing classification. Firstly, for the inputs of the proposed network, the interactive data fusion strategy (IDFS) combines generalized-intensity-hue-saturation (GIHS), and discrete wavelet transform (DWT) to interfuse patch pairs of the PAN and the MS images, so as to increase the similarity between them, thus reduce the difference in information between them. Secondly, for the branches of feature extraction, we design an adaptive spatial attention module (ASA-Module) and an adaptive channel attention module (ACA-Module) to strengthen spatial features from both larger-sized with smaller-sized targets and spectral features among channels. Finally, we insert the ASA-Module and ACA-Module into the residual modules to form a triple-branch network and use the common spatial-channel features extracted by the Fusion_Branch to gradually enhance the pure independent features extracted by the PAN_Branch and the MS_Branch, respectively. The experimental results indicate that SCPF-ResNet can achieve competitive performance. • An interactive data fusion strategy is proposed. • Two adaptive attention modules to highlight the advantages of two different data. • A triple-branch progressive fusion residual network for classification is proposed. [ABSTRACT FROM AUTHOR]

Ship detection is a significant and challenging task in remote sensing. Due to the arbitrary-oriented property and large aspect ratio of ships, most of the existing detectors adopt rotation boxes to represent ships. However, manual-designed rotation anchors are needed in these detectors, which causes multiplied computational cost and inaccurate box regression. To address the abovementioned problems, an anchor-free rotation ship detector, named GRS-Det, is proposed, which mainly consists of a feature extraction network with selective concatenation module (SCM), a rotation Gaussian-Mask model, and a fully convolutional network-based detection module. First, a U-shape network with SCM is used to extract multiscale feature maps. With the help of SCM, the channel unbalance problem between different-level features in feature fusion is solved. Then, a rotation Gaussian-Mask is designed to model the ship based on its geometry characteristics, which aims at solving the mislabeling problem of rotation bounding boxes. Meanwhile, the Gaussian-Mask leverages context information to strengthen the perception of ships. Finally, multiscale feature maps are fed to the detection module for classification and regression of each pixel. Our proposed method, evaluated on ship detection benchmarks, including HRSC2016 and DOTA Ship data sets, achieves state-of-the-art results. [ABSTRACT FROM AUTHOR]

The content-based remote sensing image retrieval (CBRSIR) has attracted increasing attention with the number of remote sensing (RS) images growing explosively. Benefiting from the strong capacity of the deep convolutional neural network (DCNN), the performance of CBRSIR has been improved in recent years. Although great successes have been obtained, learning the RS images’ representative features and enhancing the retrieval efficiency for the large-scale CBRSIR tasks are still two challenging problems. In this article, we propose a new CBRSIR method named feature and hash (FAH) learning, which consists of a deep feature learning model (DFLM) and an adversarial hash learning model (AHLM). The DFLM aims at learning the RS images’ dense features to guarantee the retrieval precision. In the DFLM, the DCNN and the proposed feature aggregation are integrated to capture the multiscale features. Then, the discrimination of the obtained features can be highlighted by the attention map in the developed attention branch. The AHLM maps the dense features onto the compact hash codes so that the retrieval efficiency can be improved. The AHLM contains a hash learning submodel and an adversarial regularization submodel. In particular, the hash learning submodel learns the real-valued hash codes that are similarity preserved by semantic supervisions. The adversarial regularization submodel regularizes the real-valued hash codes to learn the discrete uniform distribution with possible values 0 and 1. In this way, the hash codes are coding-balanced and the quantization errors are reduced. Encouraging experimental results counted on three public benchmark data sets demonstrate that our FAH can achieve competitive performance in the CBRSIR task compared with many existing hash learning methods. [ABSTRACT FROM AUTHOR]

Tremendous progresses have been made in remote sensing image captioning (RSIC) task in recent years, yet there still some unresolved problems: (1) facing the gap between the visual features and semantic concepts, (2) reasoning the higher-level relationships between semantic concepts. In this work, we focus on injecting high-level visual-semantic interaction into RSIC model. Firstly, the semantic concept extractor (SCE), end-to-end trainable, precisely captures the semantic concepts contained in the RSIs. In particular, the visual-semantic co-attention (VSCA) is designed to grain coarse concept-related regions and region-related concepts for multi-modal interaction. Furthermore, we incorporate the two types of attentive vectors with semantic-level relational features into a consensus exploitation (CE) block for learning cross-modal consensus-aware knowledge. The experiments on three benchmark data sets show the superiority of our approach compared with the reference methods. • We propose an end-to-end trainable network architecture for remote sensing image captioning task, aiming to extract explicit high-level semantic from remote sensing image, learn discriminative cross-modal linguistic-aware features, and exploit cross-modal alignment. • To capture the explicit semantic concepts, semantic concept extractor is designed to generate a series of semantic words. The designed network based on visual features and semantic concepts is effective for visual-semantic features extracting. • For modeling the visual features and semantic concepts, a visual-semantic co-attention module with two parallel attentive branches is designed, which learns the linguistic awareness between visual features and enhances the region-related semantic contents. • In order to facilitate the alignment between cross-modal features, a consensus exploitation is devised for exploring the multi-modal consensus-aware representations based on the constructed concept correlation graph. [ABSTRACT FROM AUTHOR]

The task of image captioning involves the generation of a sentence that can describe an image appropriately, which is the intersection of computer vision and natural language. Although the research on remote sensing image captions has just started, it has great significance. The attention mechanism is inspired by the way humans think, which is widely used in remote sensing image caption tasks. However, the attention mechanism currently used in this task is mainly aimed at images, which is too simple to express such a complex task well. Therefore, in this paper, we propose a multi-level attention model, which is a closer imitation of attention mechanisms of human beings. This model contains three attention structures, which represent the attention to different areas of the image, the attention to different words, and the attention to vision and semantics. Experiments show that our model has achieved better results than before, which is currently state-of-the-art. In addition, the existing datasets for remote sensing image captioning contain a large number of errors. Therefore, in this paper, a lot of work has been done to modify the existing datasets in order to promote the research of remote sensing image captioning. [ABSTRACT FROM AUTHOR]

Semantic segmentation of high-resolution remote sensing images is highly challenging due to the presence of a complicated background, irregular target shapes, and similarities in the appearance of multiple target categories. Most of the existing segmentation methods that rely only on simple fusion of the extracted multi-scale features often fail to provide satisfactory results when there is a large difference in the target sizes. Handling this problem through multi-scale context extraction and efficient fusion of multi-scale features, in this paper we present an end-to-end multi-scale adaptive feature fusion network (MANet) for semantic segmentation in remote sensing images. It is a coding and decoding structure that includes a multi-scale context extraction module (MCM) and an adaptive fusion module (AFM). The MCM employs two layers of atrous convolutions with different dilatation rates and global average pooling to extract context information at multiple scales in parallel. MANet embeds the channel attention mechanism to fuse semantic features. The high- and low-level semantic information are concatenated to generate global features via global average pooling. These global features are used as channel weights to acquire adaptive weight information of each channel by the fully connected layer. To accomplish an efficient fusion, these tuned weights are applied to the fused features. Performance of the proposed method has been evaluated by comparing it with six other state-of-the-art networks: fully convolutional networks (FCN), U-net, UZ1, Light-weight RefineNet, DeepLabv3+, and APPD. Experiments performed using the publicly available Potsdam and Vaihingen datasets show that the proposed MANet significantly outperforms the other existing networks, with overall accuracy reaching 89.4% and 88.2%, respectively and with average of F1 reaching 90.4% and 86.7% respectively. [ABSTRACT FROM AUTHOR]

Polarimetric synthetic aperture radar (PolSAR) image classification is a recent technology with great practical value in the field of remote sensing. However, due to the time-consuming and labor-intensive data collection, there are few labeled datasets available. Furthermore, most available state-of-the-art classification methods heavily suffer from the speckle noise. To solve these problems, in this paper, a novel semi-supervised algorithm based on self-training and superpixels is proposed. First, the Pauli-RGB image is over-segmented into superpixels to obtain a large number of homogeneous areas. Then, features that can mitigate the effects of the speckle noise are obtained using spatial weighting in the same superpixel. Next, the training set is expanded iteratively utilizing a semi-supervised unlabeled sample selection strategy that elaborately makes use of spatial relations provided by superpixels. In addition, a stacked sparse auto-encoder is self-trained using the expanded training set to obtain classification results. Experiments on two typical PolSAR datasets verified its capability of suppressing the speckle noise and showed excellent classification performance with limited labeled data. [ABSTRACT FROM AUTHOR]