Search

Several deep learning and transformer models have been recommended in previous research to deal with the classification of hyperspectral images (HSIs). Among them, one of the most innovative is the bidirectional encoder representation from transformers (BERT), which applies a distance-independent approach to capture the global dependency among all pixels in a selected region. However, this model does not consider the local spatial-spectral and spectral sequential relations. In this paper, a dual-dimensional (i.e., spatial and spectral) BERT (the so-called D2BERT) is proposed, which improves the existing BERT model by capturing more global and local dependencies between sequential spectral bands regardless of distance. In the proposed model, two BERT branches work in parallel to investigate relations among pixels and spectral bands, respectively. In addition, the layer intermediate information is used for supervision during the training phase to enhance the performance. We used two widely employed datasets for our experimental analysis. The proposed D2BERT shows superior classification accuracy and computational efficiency with respect to some state-of-the-art neural networks and the previously developed BERT model for this task. [ABSTRACT FROM AUTHOR]

One of the challenges in the field of remote sensing is how to automatically identify and classify high-resolution remote sensing images. A number of approaches have been proposed. Among them, the methods based on low-level visual features and middle-level visual features have limitations. Therefore, this paper adopts the method of deep learning to classify scenes of high-resolution remote sensing images to learn semantic information. Most of the existing methods of convolutional neural networks are based on the existing model using transfer learning, while there are relatively few articles about designing of new convolutional neural networks based on the existing high-resolution remote sensing image datasets. In this context, this paper proposes a multi-view scaling strategy, a new convolutional neural network based on residual blocks and fusing strategy of pooling layer maps, and uses optimization methods to make the convolutional neural network named RFPNet more robust. Experiments on two benchmark remote sensing image datasets have been conducted. On the UC Merced dataset, the test accuracy, precision, recall, and F1-score all exceed 93%. On the SIRI-WHU dataset, the test accuracy, precision, recall, and F1-score all exceed 91%. Compared with the existing methods, such as the most traditional methods and some deep learning methods for scene classification of high-resolution remote sensing images, the proposed method has higher accuracy and robustness. [ABSTRACT FROM AUTHOR]

Research on transformers in remote sensing (RS), which started to increase after 2021, is facing the problem of a relative lack of review. To understand the trends of transformers in RS, we undertook a quantitative analysis of the major research on transformers over the past two years by dividing the application of transformers into eight domains: land use/land cover (LULC) classification, segmentation, fusion, change detection, object detection, object recognition, registration, and others. Quantitative results show that transformers achieve a higher accuracy in LULC classification and fusion, with more stable performance in segmentation and object detection. Combining the analysis results on LULC classification and segmentation, we have found that transformers need more parameters than convolutional neural networks (CNNs). Additionally, further research is also needed regarding inference speed to improve transformers' performance. It was determined that the most common application scenes for transformers in our database are urban, farmland, and water bodies. We also found that transformers are employed in the natural sciences such as agriculture and environmental protection rather than the humanities or economics. Finally, this work summarizes the analysis results of transformers in remote sensing obtained during the research process and provides a perspective on future directions of development. [ABSTRACT FROM AUTHOR]

Satellite images have drawn increasing interest from a wide variety of users, including business and government, ever since their increased usage in important fields ranging from weather, forestry and agriculture to surface changes and biodiversity monitoring. Recent updates in the field have also introduced various deep learning (DL) architectures to satellite imagery as a means of extracting useful information. However, this new approach comes with its own issues, including the fact that many users utilize ready-made cloud services (both public and private) in order to take advantage of built-in DL algorithms and thus avoid the complexity of developing their own DL architectures. However, this presents new challenges to protecting data against unauthorized access, mining and usage of sensitive information extracted from that data. Therefore, new privacy concerns regarding sensitive data in satellite images have arisen. This research proposes an efficient approach that takes advantage of privacy-preserving deep learning (PPDL)-based techniques to address privacy concerns regarding data from satellite images when applying public DL models. In this paper, we proposed a partially homomorphic encryption scheme (a Paillier scheme), which enables processing of confidential information without exposure of the underlying data. Our method achieves robust results when applied to a custom convolutional neural network (CNN) as well as to existing transfer learning methods. The proposed encryption scheme also allows for training CNN models on encrypted data directly, which requires lower computational overhead. Our experiments have been performed on a real-world dataset covering several regions across Saudi Arabia. The results demonstrate that our CNN-based models were able to retain data utility while maintaining data privacy. Security parameters such as correlation coefficient (−0.004), entropy (7.95), energy (0.01), contrast (10.57), number of pixel change rate (4.86), unified average change intensity (33.66), and more are in favor of our proposed encryption scheme. To the best of our knowledge, this research is also one of the first studies that applies PPDL-based techniques to satellite image data in any capacity. [ABSTRACT FROM AUTHOR]

High-Performance Computing (HPC) has recently been attracting more attention in remote sensing applications due to the challenges posed by the increased amount of open data that are produced daily by Earth Observation (EO) programs. The unique parallel computing environments and programming techniques that are integrated in HPC systems are able to solve large-scale problems such as the training of classification algorithms with large amounts of Remote Sensing (RS) data. This paper shows that the training of state-of-the-art deep Convolutional Neural Networks (CNNs) can be efficiently performed in distributed fashion using parallel implementation techniques on HPC machines containing a large number of Graphics Processing Units (GPUs). The experimental results confirm that distributed training can drastically reduce the amount of time needed to perform full training, resulting in near linear scaling without loss of test accuracy. [ABSTRACT FROM AUTHOR]

In various applications of airborne laser scanning (ALS), the classification of the point cloud is a basic and key step. It requires assigning category labels to each point, such as ground, building or vegetation. Convolutional neural networks have achieved great success in image classification and semantic segmentation, but they cannot be directly applied to point cloud classification because of the disordered and unstructured characteristics of point clouds. In this paper, we design a novel convolution operator to extract local features directly from unstructured points. Based on this convolution operator, we define the convolution layer, construct a convolution neural network to learn multi-level features from the point cloud, and obtain the category label of each point in an end-to-end manner. The proposed method is evaluated on two ALS datasets: the International Society for Photogrammetry and Remote Sensing (ISPRS) Vaihingen 3D Labeling benchmark and the 2019 IEEE Geoscience and Remote Sensing Society (GRSS) Data Fusion Contest (DFC) 3D dataset. The results show that our method achieves state-of-the-art performance for ALS point cloud classification, especially for the larger dataset DFC: we get an overall accuracy of 97.74% and a mean intersection over union (mIoU) of 0.9202, ranking in first place on the contest website. [ABSTRACT FROM AUTHOR]

In recent years, hyperspectral images (HSIs) have attained considerable attention in computer vision (CV) due to their wide utility in remote sensing. Unlike images with three or lesser channels, HSIs have a large number of spectral bands. Recent works demonstrate the use of modern deep learning based CV techniques like convolutional neural networks (CNNs) for analyzing HSI. CNNs have receptive fields (RFs) fueled by learnable weights, which are trained to extract useful features from images. In this work, a novel multi-receptive CNN module called GhoMR is proposed for HSI classification. GhoMR utilizes blocks containing several RFs, extracting features in a residual fashion. Each RF extracts features which are used by other RFs to extract more complex features in a hierarchical manner. However, the higher the number of RFs, the greater the associated weights, thus heavier is the network. Most complex architectures suffer from this shortcoming. To tackle this, the recently found Ghost module is used as the basic building unit. Ghost modules address the feature redundancy in CNNs by extracting only limited features and performing cheap transformations on them, thus reducing the overall parameters in the network. To test the discriminative potential of GhoMR, a simple network called GhoMR-Net is constructed using GhoMR modules, and experiments are performed on three public HSI data sets—Indian Pines, University of Pavia, and Salinas Scene. The classification performance is measured using three metrics—overall accuracy (OA), Kappa coefficient (Kappa), and average accuracy (AA). Comparisons with ten state-of-the-art architectures are shown to demonstrate the effectiveness of the method further. Although lightweight, the proposed GhoMR-Net provides comparable or better performance than other networks. The PyTorch code for this study is made available at the iamarijit/GhoMR GitHub repository. [ABSTRACT FROM AUTHOR]