Few-shot learning has achieved great success in computer vision. However, when applied to Synthetic Aperture Radar Automatic Target Recognition (SAR-ATR), it tends to demonstrate a bad performance due to the ignorance of the differences between SAR images and optical ones. What is more, the same transformation on both images may cause different results, even some unexpected noise. In this paper, we propose an improved Prototypical Network (PN) based on Spatial Transformation, also known as ST-PN. Cascaded after the last convolutional layer, a spatial transformer module implements a feature-wise alignment rather than a pixel-wise one, so more semantic information can be exploited. In addition, there is always a huge divergence even for the same target when it comes to pixel-wise alignment. Moreover, it reduces computational cost with fewer parameters of the deeper layer. Here, a rotation transformation is used to reduce the discrepancies caused by different observation angles of the same class. Thefinal comparison of four extra losses indicates that a single cross-entropy loss is good enough to calculate the loss of distances. Our work achieves state-of-the-art performance on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset.

Few-shot synthetic aperture radar automatic target recognition (SAR-ATR) aims to recognize the targets of the images (query images) based on a few annotated images (support images). Such a task requires modeling the relationship between the query and support images. In this paper, we propose the instance-aware transformer (IAT) model. The IAT exploits the power of all instances by constructing the attention map based on the similarities between the query feature and all support features. The query feature aggregates the support features based on the attention values. To align the features of the query and support images in IAT, the shared cross-transformer keep all the projections in the module shared across all features. Instance cosine distance is used in training to minimize the distance between the query feature and the support features. In testing, to fuse the support features of the same class into the class representation, Euclidean (Cosine) Loss is used to calculate the query-class distances. Experiments on the two proposed few-shot SAR-ATR test sets based on MSTAR demonstrate the superiority of the proposed method.

Pansharpening aims to generate a high-resolution multi-spectral (HR-MS) image given a paired panchromatic (PAN) image and low-resolution multi-spectral (LR-MS) image. Though existing pansharpening methods have made remarkable progress, the fusion pipeline does not fully adapt to the distinct characteristics of the PAN and LR-MS images. In this paper, to fully exploit the complementary modality of the two images, we propose a novel and efficient asymmetric bidirectional fusion network (ABFNet). The ABFNet consists of the two customized fusion modules with asymmetric architectures, which aim to reinforce the PAN and LR-MS images respectively. Specifically, the spectral colorization module recalibrates the scale and bias of the PAN features using weights generated by the LR-MS features, which aims to inject spectral information into the PAN features without breaking their spatial continuity. To transfer spatial details from the PAN features into the LR-MS features, the spatial restoration codebook module refines the LR-MS features with point-to-point restoration codebooks learned from the PAN features. By incorporating the two modules in multiple stages, ABFNet enjoys a high capability for capturing both spectral and spatial dependencies. Extensive experiments over multiple satellite datasets demonstrate the effectiveness of the proposed methods.

The semantic segmentation of remote sensing images requires distinguishing local regions of different classes and exploiting a uniform global representation of the same-class instances. Such requirements make it necessary for the segmentation methods to extract discriminative local features between different classes and to explore representative features for all instances of a given class. While common deep convolutional neural networks (DCNNs) can effectively focus on local features, they are limited by their receptive field to obtain consistent global information. In this paper, we propose a memory-augmented transformer (MAT) to effectively model both the local and global information. The feature extraction pipeline of the MAT is split into a memory-based global relationship guidance module and a local feature extraction module. The local feature extraction module mainly consists of a transformer, which is used to extract features from the input images. The global relationship guidance module maintains a memory bank for the consistent encoding of the global information. Global guidance is performed by memory interaction. Bidirectional information flow between the global and local branches is conducted by a memory-query module, as well as a memory-update module, respectively. Experiment results on the ISPRS Potsdam and ISPRS Vaihingen datasets demonstrated that our method can perform competitively with state-of-the-art methods.

Remote sensing images play a crucial role in remote sensing target detection and 3D remote sensing modeling, and the enhancement of resolution holds significant application implications. The task of remote sensing target detection requires a substantial amount of high-resolution remote sensing images, while 3D reconstruction tasks generate denser models from diverse view perspectives. However, high-resolution remote sensing images are often limited due to their high acquisition costs, a scarcity of acquisition views, and restricted view perspective variations, which pose challenges for remote sensing tasks. In this paper, we propose an advanced method for a high-resolution remote sensing novel view synthesis by integrating attention mechanisms with neural radiance fields to address the scarcity of high-resolution remote sensing images. To enhance the relationships between sampled points and rays and to improve the 3D implicit model representation capability of the network, we introduce a point attention module and batch attention module into the proposed framework. Additionally, a frequency-weighted position encoding strategy is proposed to determine the significance of each frequency for position encoding. The proposed method is evaluated on the LEVIR-NVS dataset and demonstrates superior performance in quality assessment metrics and visual effects compared to baseline NeRF (Neural Radiance Fields) and ImMPI (Implicit Multi-plane Images). Overall, this work presents a promising approach for a remote sensing novel view synthesis by leveraging attention mechanisms and frequency-weighted position encoding.

Pansharpening aims to reconstruct the high-resolution multi-spectral image (HR-MS) by fusing the low-resolution multi-spectral image (LR-MS) and the panchromatic images (PAN). The effectiveness of pansharpening algorithms highly relies on the fusion mechanism of two images with different modality. In this paper, we propose a novel Progressive Reconstruction Network (PRNet) with adaptive frequency adjustment to address the spatial and frequency discrepancies between LR-MS and PAN images. Aiming to simultaneously capturing contextual information and frequency characteristics, our PRNet fuses the LR-MS and PAN features in bi-scale frequency domains. Specifically, we use the wavelet transform to decompose the PAN features into bi-scale coefficients, dubbed as coarse-scale and fine-scale coefficients, respectively. Initially, the LR-MS features are fused into the coarse-scale coefficients first, then the coarse-scale fused results are used to refine the fine-scale coefficients, leading to the progressive reconstruction mechanism. Additionally, to mitigate aliasing between LR-MS features and PAN coefficients across multiple frequency domains, we propose the adaptively frequency adjustment module to control the fusion weights and recalibrate the frequency responses of the reconstruct features. Experimental evaluations on reduced-resolution and full-resolution pansharpening datasets demonstrate that our method outperforms state-of-the-art pansharpening methods quantitatively and qualitatively, affirming its effectiveness. The code will be released soon.