Synthetic Aperture Radar (SAR) images have become a primary data source for change detection due to their all-weather, all-day imaging capability, high resolution, and strong sensitivity to ground surface variations. However, the complex scattering characteristics of SAR make distinguishing between changed and unchanged areas particularly challenging. Additionally, the long-tail data distribution, where changed areas constitute only a small portion of the dataset, further exacerbates the difficulty of change detection. To address these challenges, we propose a Fine-Grained Meta-Learning with Semantic Augmentation method for SAR image change detection. First, we propose a fine-grained classification strategy based on edge detection to construct a small, balanced dataset for training the meta-learner and a dataset with reduced imbalance for training the backbone classifier. This strategy enhances the feature learning capability of hard-to-classify samples and reduces the adverse effects of data imbalance. Next, we design a lightweight yet effective Multi-Layer Perceptron-based meta-learner and incorporate semantic data augmentation. The meta-learner automatically augments minority classes along meaningful semantic directions by learning appropriate class-wise covariance matrices, thereby improving the detection performance of the backbone classifier. We evaluate our method on four SAR datasets through cross-dataset experiments, demonstrating its superiority over four methods in effectiveness and robustness. Imbalance analysis further reveals that as the imbalance ratio increases, the performance of comparison methods degrades significantly, whereas our method exhibits the least performance deterioration, confirming its competent capability in handling imbalanced data.

With the development of multi-modal technology, hyperspectral image (HSI) and light detection and ranging (LiDAR) data has achieved remarkable results in land use and land cover (LULC) classification. Recently, many deep learning based feature extraction and fusion methods have improved the classification performance of LULC tasks. However, most of these methods use a single feature extractor and do not fully utilize the information of HSI and LiDAR data. Moreover, directly fusing various features obtained from feature extractor can lead to feature redundancy, resulting in model overfitting. In this paper, we develop a three-branch excitation network, named TBENet. The three branches extract spectral features, spatial features and elevation features respectively. And an excitation block is used to reduce feature redundancy and improve the generalization of the model. Contrast experiments on Houston dataset show that our proposed method outperforms other state-of-the-art methods, and ablation experiments demonstrate the effectiveness of each block.

Convolutional Neural Networks (CNNs) have achieved remarkable performance in remote sensing image classification tasks. To address the issue of high model complexity, we propose a block-level pruning strategy based on the semantic similarity analysis that no fine-tuning is required during the pruning process. By employing this strategy, we effectively reduce the complexity of the model. Furthermore, to restore the overall performance of the pruned model, we propose a teacher-student collaborative distillation strategy that enables knowledge transfer through the collaboration of the original model and the dropped-blocks model to promote the performance of the compact pruned model. Experimental results demonstrate that our pruning and distillation strategies outperform other approaches, thereby achieving favorable performance while reducing model complexity.

The research about potential landslide hazards is of great significance for early detection and prevention of geological disasters. However, the existing datasets for the landslide detection visual mainly focus on historical landslides that have already occurred. In this paper, we construct a Ningxia potential landslide hazard benchmark dataset named NXPLH for geological disaster detection of remote sensing. Specifically, it contains 315 potential landslide hazards and 1039 without landslide samples. We have released the dataset at https://github.com/Dataset-RFGroup/NXPLH-Dataset.

The landslide detection methods using remote sensing images are mostly based on the traditional convolutional neural network model with high depth and complexity. The paper proposes a lightweight method based on Depth Separable Convolution and Double Self-Attention Mechanism (DSC-DSAM) for detecting landslides in remote sensing images. This method aims to reduce storage space and improve detection speed while maintaining accuracy. In our model, it starts with using a lightweight convolutional neural network model. Then, the dual self-attention mechanism is applied to improve the accuracy. The proposed method is compared with other existing classification models, and it is shown to have advantages in memory space and detection speed while maintaining accuracy.

Bi-temporal semantic change detection(SCD) is more sophisticated than binary change detection and it provides more detailed changing information with categories. Naturally, it is more challenging than traditional binary change detection. In this paper, a Siamese CNN is proposed for SCD. For the problems of complex backgrounds of remote sensing images, we use multiscale context information and correlation to enhance SCD performance. For the problem of insufficient feature utilization between subtasks, a channel fusion module is proposed to explore the temporal correlation between bi-temporal images, which benefits the extraction of the final changing map. The experiments in this paper are conducted on the SECOND dataset. Our proposed method outperforms compared methods and obtains more completed changing maps than other methods.

Flood disasters last for a long time and are destructive, so it is necessary to obtain the submerged area in a timely and effective manner, which is very important for reducing disaster losses and monitoring floods. The main contribution of this paper is to construct a dataset for training and validation of deep learning algorithms for flood detection for Gaofen-3. To overcome the scarcity of SAR datasets for water segmentation, this paper constructs a refined water segmentation dataset named Dali Water Segmentation (Dali-WS) based on the Gaofen-3 satellite in China. The dataset provides abundant rural waters in Dali County, and 1776 chips were hand-labeled for further research. We also report extensive performance for the state-of-the-art segmentation algorithms. Additionally, comprehensive evaluations of state-of-the-art segmentation algorithms are presented, demonstrating the challenging nature of the dataset and its potential for driving further advancements in flood detection. The findings of this study are expected to contribute to the progress of flood detection and recognition research.

High-Efficient landslide recognition on remote sensing images is of great importance to hazard monitoring. In this paper, we introduce MobileL-K, a light Convolutional neural networks(CNNs) to achieve highly efficient landslide recognition. In this network, the depthwise separable convolutions with a large kernel is borrowed to exploit global features on an image. Moreover, an improved EC-based network pruning method was proposed based on continual masking. We prune the MobileL-K to apply to landslide recognition. The experiment results on a benchmark dataset show that proposed method outperforms other compared methods with smaller model size, less FLOPs and higher running speed on GPU.

It is challenging to perform robust road detection on remote sensing images in a complex scene with occlusions by plants and buildings. In this paper, an elaborate dual-decoded U-Net combined with atrous spatial pyramid pooling is proposed to tackle this scenario. In the proposed network, a dual-decoder structure is designed, where a small decoder aims to extract the attention information and it is delivered to the other decoder to enhance the context. Finally, the proposed method is verified on the DeepGlobe dataset. The experiment results demonstrate that the proposed method outperforms other compared methods.

Water area segmentation in remote sensing images is of great importance for flood monitoring. Convolutional neural networks have been successfully applied to various computer vision tasks. Among them, a U-shaped CNN known as U-Net achieves state-of-the-art performance on various types of image segmentation, including remote sensing images. However, there are still some difficulties in the water area segmentation of remote sensing images, such as complex backgrounds, cloud shading, and rough edges. In this work, we propose a multi-branch fusion U-Net (MFU-Net) method for water area segmentation with multi-modality remote sensing images. The experimental results showed that our MFU-Net can effectively and efficiently segment water area from Sentinel-1 and Sentinel-2 images, which F1, IoU and PA on the Sen1Floods11 dataset are 91.462%, 84.598% and 98.123%, respectively.

The graph neural network (GNN) has been widely applied to image analysis and recognition. Recently, a semisupervised graph convolutional network (ssGCN) method has been proposed to change detection and obtains promising performance on very-high-resolution remote sensing images. However, a synthetic aperture radar (SAR) image is subject to speckle noise, and there is no explicit structure. In this letter, an end-to-end dynamic graph-level neural network (DGLNN) is proposed to exploit the local structure of each pixel neighborhood block at a graph level and learn a more discriminative graph for change detection. Moreover, in the training of DGLNN, a$K$-nearest neighborhood is employed to reconstruct edges between nodes instead of the fixed edges between two nodes so that each node exploits the features from different neighbor nodes. The proposed method is verified by cross-domain SAR image change detection on four sets of SAR images and compared with five state-of-the-art deep-learning-based SAR image change detection methods. The overall experimental results show that the proposed DGLNN obtains outstanding performance.

In synthetic aperture radar (SAR) image change detection, it is quite challenging to exploit the changing information from the noisy difference image subject to the speckle. Although convolutional neural network has been proposed for feature learning, it is necessary to collect numerous of samples to train a perfect model, which is difficult to achieve. In this paper, we propose a few-shot learning-based neural network to exploit the changed information from the noisy difference image. Being different from traditional training method with numerous labeled samples, in the proposed method, fewer samples are used to train a neural network. Finally, we verify our proposed method on four challenging datasets of bitemporal SAR images. Experimental results demonstrate that the difference map obtained by our proposed method outperforms than other state-of-art methods.

Graph neural network (GNN) has been widely applied to computer vision as well as remote sensing image analysis. In this paper, we propose an end-to-end graph-level neural network (GLNN) for SAR image change detection. In the proposed method, a GNN is applied to exploit the local structure of an image patch at a graph-level and learn a more discriminative representation. Then, based on these graph representations, change detection is conducted by training an end-to-end neural network. Our method is verified on four cross-dataset of SAR image and compared with three state-of-art deep learning SAR image change detection methods. The experimental results show that the proposed GLNN outperforms other compared methods.

Locoregional recurrence (LRR) remains one of leading causes in head and neck (H&N) cancer treatment failure despite the advancement of multidisciplinary management. Accurately predicting LRR in early stage can help physicians make an optimal personalized treatment strategy. In this study, we propose an end-to-end multi-modality and multi-view convolutional neural network model (mMmV-CNN) for LRR prediction in H&N cancer. In mMmV, a dimension reduction operator is designed, projecting the 3D volume onto 2D images in different directions, and a multi-view strategy is used to replace the original 3D method, which reduces the complexity of the algorithm while preserving important 3D information. Meanwhile, multi-modal data is used for the classification by making full use of the complementary information from cross modality data. Furthermore, we design a multi-modality deep neural network which is trained in an end-to-end manner and jointly optimize the deep features of CT, PET and clinical features. A H&N dataset which consists of 206 patients was used to evaluate the performance. Experimental results demonstrated that mMm V-CNN can obtain an AUC value of 0.81 and outperform a state of the art CNN-based method.

Recently, many convolution neural networks have been successfully employed in bitemporal SAR image change detection. However, most of those networks are too heavy where large memory are necessary for storage and calculation. To reduce the computational and spatial complexity and facilitate the change detection on edge devices, in this paper, we propose a lightweight neural network for bitemporal SAR image change detection. In the proposed network, we replace the regular convolutional layers with bottlenecks, which will not increase the number of channels. Furthermore, we employ dilated convolutional kernels with a few non-zero entries which reduces the FLOPs in convlutional operators. Comparing with traditional neural network, our lightweight neural network will be faster, less FLOPs and parameters. We verify our lightweight neural network on two sets of bitemporal SAR images. The experimental results show that the proposed network can obtain the comparable performance with those heavy-weight neural network.

In synthetic aperture radar (SAR) image change detection, it is quite challenging to exploit the changing information from the noisy difference image subject to the speckle. In this paper, we propose a novel mutli-scale average pooling (MSAP) network to exploit the changed information from the noisy difference image. Being different from traditional convolutional network with only an one-scale pooling kernel, in the proposed method, multi -scale pooling kernels are equipped in convolutional network to obtain the spatial context information on changed regions from the difference image. Finally, we verify our proposed method on four challenging datasets of bitemporal SAR images. Experimental results demonstrate that the difference map obtained by our proposed method outperforms than other state-of-art methods.

Recently, many convolution neural networks have been successfully employed in bitemporal SAR image change detection. However, most methods are developed based on the traditional framework that exploits the changed region from a difference image (DI) that is usually subject to the speckle. To essentially solve this issue, in this paper, we propose a deep canonic correlation network for bitemporal SAR image. In the proposed network, bitemporal SAR images and its corresponding DI are taken as the inputs and then three deep neural networks are designed to employ their features, respectively. Then the changed regions are obtained by the exploited features. Finally, we compare the proposed method with other deep learning methods and perform the comparison on four sets of bitemporal SAR images. The experimental results show that our proposed method outperforms other methods.

The log-ratio (LR) operator has been widely employed to generate the difference image for synthetic aperture radar (SAR) image change detection. However, the difference image generated by this pixelwise operator can be subject to SAR images speckle and unavoidable registration errors between bitemporal SAR images. In this letter, we proposed a spatial metric learning method to obtain a difference image that is more robust to the speckle by learning a metric from a set of constraint pairs. In the proposed method, the spatial context is considered in constructing constraint pairs, each of which consists of patches in the same location of bitemporal SAR images. Then, a semidefinite positive metric matrix M can be obtained by the optimization with the max-margin criterion. Finally, we verify our proposed method on four challenging data sets of bitemporal SAR images. Experimental results demonstrate that the difference map obtained by our proposed method outperforms than other state-of-the-art methods.

Change detection is a quite challenging task due to the imbalance between unchanged and changed class. In addition, the traditional difference map generated by log-ratio is subject to the speckle, which will reduce the accuracy. In this letter, an imbalanced learning-based change detection is proposed based on PCA network (PCA-Net), where a supervised PCA-Net is designed to obtain the robust features directly from given multitemporal synthetic aperture radar (SAR) images instead of a difference map. Furthermore, to tackle with the imbalance between changed and unchanged classes, we propose a morphologically supervised learning method, where the knowledge in the pixels near the boundary between two classes is exploited to guide network training. Finally, our proposed PCA-Net can be trained by the data sets with available reference maps and applied to a new data set, which is quite practical in change detection projects. Our proposed method is verified on five sets of multiple temporal SAR images. It is demonstrated from the experiment results that with the knowledge in training samples from the boundary, the learned features benefit change detection and make the proposed method outperform than supervised methods trained by randomly drawing samples.

In this paper, a novel stacked marginal discriminative autoencoder (SMDAE) method is proposed for hyperspectral image classification. It uses a deep neural network to learn discriminative features from hyperspectral images automatically. In hyperspectral images, the collection of training samples is difficult. When the number of training samples is not enough, these training samples are difficult to estimate the statistical distribution of hyperspectral images accurately. In order to solve the small sample problem and improve the classification performance of the autoencoder, the marginal samples are selected through the distribution characteristics of samples. The marginal samples are searched based on k nearest neighbors between different classes. These samples are used to fine-tune the SMDAE network. The experimental results show that the proposed SMDAE method can achieve satisfying performance under small training set.