Panchromatic images (PANs) and multispectral (MS) images (MSs) are widely used for dual-source remote sensing image classification, gradually becoming a research hotspot. However, making the most of dual-source image information with insufficiently labeled samples is a significant challenge. This article proposes a progressive semi-distillation model (PSDM) to classify dual-source remote sensing images with insufficient samples. We design a framework of rookie teacher network (RTN)-teaching assistant system (TAS)-student grouping network (SGN) in the case of a traditional teacher network (TN) (i.e., rookie TN (RTN)) that does not provide excellent guidance to student network (SN) due to insufficient samples. The PSDM expands the samples and compresses the space through the RTN-SGN structure to cope with the dilemma of insufficient samples. To make RTN better guide the SGN, we design TAS, which can gradually guide SGN to learn the samples from easy to difficult. It can also further assist SGN training to improve the classification performance of SGN with insufficient samples. We design SGN and add cooperation and CorM to better learn dual-source information. These strategies can eliminate SGN’s over-dependence on the RTN, help SGN outperform the RTN, and achieve the effect of semi-distillation. Experimental results and theoretical analysis have sufficiently pointed out the proposed method’s accuracy, efficiency, and robustness under insufficient sample situations. Our model is available at https://github.com/MarjordCpz/PSDM.

Dual-modal remote sensing classification combines different modal images effectively by using internal consistency. The representation of dual-modal remote sensing images may differ, but they share a common underlying characteristic. Deep learning networks are required to preserve relevant information while removing redundant information. On the one hand, it is necessary to explore the inherent commonality among different modalities, reduce the differences between modalities, and bridge the high-level semantic gap across features, thus creating fusion conditions. On the other hand, during feature fusion, it is essential to emphasize the complementary advantages of each modality, highlighting their unique features and removing the shared redundant features. This paper proposes a dynamic quadruplet transition learning method for dual-modal image classification. Compared to the previous methods, we reduce the artificial design of the network, following the original data’s drive, and achieve good performance. Specifically, we acquire the auxiliary modalities through modal transition learning, which constitute quadruplets. The auxiliary modalities contain unique features from the original modalities, which makes them serve as bridges between the original modalities. We capture the differences between feature distributions in the quadruplets through feature extraction net to balance and narrow the semantic gap between different modalities. The auxiliary modalities can more accurately capture the unique features of different modalities and achieve a more hierarchical feature fusion instead of a violent direct fusion. Finally, we introduce quadruplets measures to screen out better quadruplets during training. The results on multiple datasets have verified the effectiveness and stability of the proposed algorithm. Our code and model are available at https://github.com/Xidian-AIGroup190726/RS-classification-DQTL.

With the rapid development of earth observation technology, the joint classification of panchromatic (PAN) and multispectral (MS) images has gained significant research value. However, despite the acquisition of large amounts of data, the few-shot problem often arises due to insufficient labeled data, and the function space of the network degrades the generalization performance. In this paper, we propose a recurrent progressive few-shot network (RPF-Net) for the classification of dual-source remote sensing images. It mainly consists of two parts: solving for the optimized fusion direction and adaptive feature-trusted decision-level fusion. In the first part, considering the problem of insufficient function space constraints under few-shot conditions, we propose representative-reinforcement learning, which performs the next fusion step by analyzing the state of the current moment and selecting the optimal action. This recurrent progressive propagation process dynamically adjusts the fusion features, guiding them toward the optimal fusion direction within a larger function space under few-shot conditions. In the second part, considering that the importance of different source features in multiple fusions is different, we focus on uncertainty theory and perform focused decision-level fusion by analyzing the characteristics of different source features. This network can dynamically adjust the fusion direction and fusion method of features, solving the problem of too large function space under few-shot conditions. The results on multiple datasets have verified the effectiveness and stability of the proposed algorithm. Our code is available at: https://github.com/cominclip/RPF-Net.

Pan-sharpening refers to fusing remote sensing multispectral (MS) and panchromatic (PAN) images to generate high-resolution multispectral (HR-MS) images. Recent advancements in deep learning-based pan-sharpening techniques have shown promising results. However, they face the following two issues. On one hand, there is a modality gap between MS and PAN images. Directly fusing them can lead to spectral and spatial distortions. On the other hand, the fusion process is prone to information loss, which can lead to image blurriness. To tackle these issues, we develop a Transformer-based model: FAFormer, which incorporates frequency analysis and focuses on the correlation and specificity of the PAN and MS images. Focusing on correlation can reduce the spectral and spatial distortions while focusing on specificity can reflect the specific information from MS and PAN images in the fusion result. We utilize the Discrete Wavelet Transform (DWT) to obtain the correlate and specific features. We introduce bijective functions based on the Transformer to design an Integrated Attention Block (IAB). As a critical component of the model, it effectively utilizes the correlation and specificity of the two images. In designing the model’s overall framework, we employ a Correlative Feature Attention Module (CFAM) to leverage the correlation between MS and PAN. We utilize a Specific Feature Attention Module (SFAM) to integrate specific information into fused features gradually. Experimental results show that our method improves pan-sharpening performance and has practical value. Codes are available at https://github.com/Xidian-AIGroup190726/FAFormer.

Pan-sharpening is the process of fusing panchromatic (PAN) and multispectral (MS) images. Its critical focus lies in accurately capturing the contour information from the PAN image during the fusion process and presenting it at a high resolution. However, existing deep learning methods lack the precise capture of delicate and smooth contour information, resulting in contour diffusion that affects the fusion results. Therefore, we introduce contourlet decomposition to capture multiscale directional delicate contour features and construct multiscale graph structures for semantic mining of dual-source contour features, continually updated through dynamic learning. By incorporating global features, we guide the multihead attention mechanism with directional decoding, enabling the network to pay more attention to high-resolution contour features, thereby gaining an advantage in image reconstruction. Cross-decoding between modalities provides strong representational capabilities for the advantageous features of both modalities, effectively enhancing the sharpening effect. Our algorithm achieves state-of-the-art results, and its effectiveness and advantages have been thoroughly validated across multiple datasets, including GaoFen-2, WorldView2, WorldView3, etc. Our code is available at https://github.com/Xidian-AIGroup190726/CDFInet.

In recent years, object detection in Synthetic-Aperture Radar (SAR) images still faces many challenges, especially for ship detection. Small or dense ships are vulnerable to the interference of complex scenes such as ports and land. In feature extraction, a large amount of redundant information on the feature map will further reduce the network’s attention to small-sized ships. Therefore, in this paper, we propose a network called neighborhood removal-and-emphasis network (NRE-Net), including an object neighborhood removal (ONR) strategy and a neighborhood feature emphasis (NFE) module. Among them, the ONR strategy directly removes the complex background information around the ship before feature extraction, only retaining effective contextual neighborhoods conducive to ship detection and avoiding the interference of complex background information on the network. The NFE module is based on ONR to extract features and form a weight map of small-sized ships or complex images. This module can adaptively recognize the detection neighborhood of each ship and highlight the detection ship on the feature map. Our network has validated the effectiveness of the method on multiple SAR ship datasets and improved the object AP for each size. Our code is available at: https://github.com/Xidian-AIGroup190726/NRENet.

In recent years, owing to the swift progression in sensor technology and the extensive utilization of remote sensing imagery, obtaining and using high-quality remote sensing images is increasingly important. Among them, it is necessary to address the classification problem of panchromatic remote sensing images and multispectral remote sensing images. In this field of research, cleverly eliminating modal differences, removing redundancy, and better integrating information has become a challenge. In this article, we propose a DBFR-AENet for the multisource remote sensing image classification task. First, the IFFS strategy aims to design different feature branches to pick up the advantageous features of multispectral and panchromatic images separately. It filters redundant information and obtains useful information with higher purity. Second, the Bi-VAE strategy aims to eliminate modal differences by constructing a low-dimensional shared space. The dual-source image is input into the encoder to obtain the latent encoding in the latent space. The goal of feature alignment can be achieved in the potential shared space. Then, perform feature fusion. Finally, classify the image after feature fusion.

As a hot research topic in remote sensing, effectively integrating the advantageous features of multispectral and panchromatic images is the main challenge for fusing these two remote sensing images. This article proposes a multiscale frequency fusion network based on ConvGRU. To address the underutilization of texture features, we extract multiscale bandpass and low-pass sub-bands representing texture and content features through Contourlet decomposition. Multiscale bandpass sub-bands contain more comprehensive and concentrated texture details. Then, by proposing a multiscale frequency feature extractor based on ConvGRU, we effectively integrate and enhance sub-bands of different scales and frequencies, fully utilizing the characteristics of multispectral and panchromatic images and scale transmission. With these enhanced sub-band features, we obtain more comprehensive scale-enhanced texture features. Simultaneously, content features are also preserved as dual-source image features. Moreover, to reduce redundancy between fused features and make more efficient use of the obtained enhanced features, we designed an Inver-band integrator (IBI) module. It can fuse enhanced features at different scales, improve the complementarity between features, and thus achieve effective fusion. Experimental results demonstrate the effectiveness and robustness of our model on multiple datasets. Our codes are available at https://github.com/Xidian-AIGroup190726/GMFnet.

In recent years, small object detection has remained challenging in remote sensing tasks. Firstly, small objects inherently have fewer pixels, making them susceptible to interference from prominently featured larger objects during feature extraction. Secondly, existing detection methods solely based on the Intersection over Union (IOU) loss are disadvantageous for small object detection and fail to leverage the rich prior information in remote sensing images. Based on these observations, we propose a significant feature elimination and sample assessment network for small object detection called SESA-Net, based on the Facet derivative model. SESA-Net introduces prior information to the network through the directional derivatives characteristic of remote sensing images. The overall network comprises the ADM module and SIA strategy. The ADM module eliminates significant responses from shallow large objects, directing the network's focus towards the features of shallow small objects. The Sample Importance Assessment (SIA) strategy addresses the limitations of the IOU loss function by using high-quality positive samples generated by ADM to provide an evaluation strategy for different positive samples of small objects. This enables the network to focus more on high-quality positive samples, thereby improving the accuracy of small object detection. The effectiveness of the proposed algorithm has been validated on multiple datasets. Our code is available at https://github.com/Xidian-AIGroup190726/RS-objectdetection-SESANet.

Recently, remote sensing technology has developed faster and faster, and obtaining high-quality panchromatic (PAN) and multispectral (MS) images has become more accessible. The complementarity between them provides new opportunities in multisource remote sensing image classification. However, solving the problem of the semantic gap between multisource high-level features and, at the same time, utilizing the complementary properties between them to reduce intersource information redundancy is still a challenge. This article constructs an (IRL)-R-3 -Net for the multisource remote sensing image classification task. Specifically, we design a cross-source interactive enhanced fusion module (CIEF-Module). For multilevel multisource features, by strengthening the dependencies of intrasource features and conducting intersource enhanced fusion, intrasource correlation features are refined, and the problem of the intersource semantic gap can be effectively alleviated. During the cross-source interaction process, we design a complementary representation supervised learning strategy (CRSL-Strategy). According to the similarities and differences of multisource features, it can adaptively promote complementary feature learning, thus generating a nonredundant multisource representation. The method has been verified to be effective on multiple RS datasets. The code is open source at: https://github.com/Xidian-AIGroup190726/Ping-Pie-I3RL-Net.git.

The rapid progress in remote sensing technology has made it convenient for satellites to capture both multispectral (MS) and panchromatic (PAN) images. MS has more spectral information, and PAN has higher spatial resolution. How to exploit the complementarity between MS and PAN images, and effectively combine their respective advantageous features while alleviating mode differences, has become a crucial research task. This article designs a style separation and mode unification network (SSMU-Net) for MS and PAN image classification from a novel and effective perspective. The network can be divided into two stages: style separation and mode unification (MU). In the style separation stage, we use wavelet decomposition and techniques similar to generative adversarial networks to preliminarily separate the information of MS and PAN into different components. These components better preserve complete information from the original data and have their own advantages in style and content. Then, we propose a symmetrical triplet traction module to perform style traction on different components, making style features more unique and content features more unified, achieving feature separation and purification. In the MU stage, we design an encoder–decoder model to reduce the impact of mode differences. The experimental results from multiple datasets validate the effectiveness of our proposed method. Our overall accuracy (OA) improved by approximately 4% on the Shanghai and Beijing datasets, and it has exceeded 99.28% on the Hohhot and Vancouver datasets. Our code is available at: https://github.com/proudpie/SSMU-Net .