Recently, ViTAE-RVSA, the first large-scale Vision Transformer (ViT) tailored for remote sensing, has demonstrated the potential of ViTs by integrating window attention with a convolutional branch. However, as this branch still adopts isotropic square kernels and conventional Feature Pyramid Networks neglect directional priors in multi-scale fusion, the resulting representations remain suboptimal for arbitrarily oriented, high-aspect-ratio, and small objects. To overcome these limitations, we propose Strip-ViTAE, a transformer-based detector that embeds a rotation-equivariant convolution (R2Conv) in the backbone to capture orientation-consistent local features. Moreover, we introduce a Parallel Strip Convolution Module (PSCM) employing horizontal and vertical strip kernels to adaptively enlarge the receptive field and strengthen local feature modeling for elongated and small objects. Finally, we design a Strip Feature Pyramid Network (StripFPN) composed of a Bottom-Up Reflow Module (Bottom-Up RM), a Strip-Based Enhancement Module (SBEM), and a Self-Attention Excitation Module (SAEM) to achieve direction-consistent cross-scale fusion. Experiments on the DOTA-v1.5 dataset show that Strip-ViTAE attains 72.82% mAP, surpassing the state-of-the-art (SOTA) by 1.35% and improving the mAP for high-aspect-ratio and small objects by 3.20%, validating the effectiveness of our method.

Fine-grained aircraft classification in optical remote sensing images relies on subtle difference of local parts. However, aircrafts in low-resolution (LR) images are depicted as small objects with blurred details and limited appearance information, resulting in the dramatic classification performance drop. We propose a novel dynamic hybrid knowledge distillation (DHKD) method to enhance fine-grained aircraft classification in LR remote sensing images. The core idea is to construct a teacher network for high-resolution (HR) images and a student network for LR images, and dynamically distill the key discriminative region attention-guided feature-based and logit-based object prior knowledge acquired from the teacher to the student to instruct its training, thereby improving the student’s fine-grained representation capability and classification performance for LR images. The distillation weights of different knowledge types are adaptive to their optimization direction similarity with that of the primary fine-grained classification task during the iterations. Extensive experiments on the MAR20 dataset demonstrate that DHKD consistently outperforms state-of-the-art distillation methods on fine-grained classification for aircrafts with absolute sizes of 32, 20, and 12 pixels in LR remote sensing images.

Tracking infrared (IR) small targets in real-time video sequences can be challenging due to the weak target information and heavy background clutter. To improve the accuracy and stability in infrared small target tracking, this letter proposes a novel method named the probability hypothesis density correlation filter (PHDCF). A stable tracking condition is introduced as a switch between the spatial–temporal regularized correlation filter (STRCF) and the derivative probability hypothesis density (PHD) filter. When the target is submerged in clutter, the PHD filter generations candidate estimates of the target according to prior tracking information, and a joint weighting by fusing the response score and motion similarity is performed to decide the new state of the target. Experimental results on real IR sequences illustrate that the PHDCF exceeds STRCF by 6.7% and 12.5% in terms of the success and precision rates, respectively, and reaches better precision and robustness than the baseline methods.

Recent deep learning based-works have made remarkable progress in Remote-Sensing Image Super-Resolution (RSISR). However, the complicated network architecture as well as a huge amount of parameters increase computational cost, hindering their practical deployment. To alleviate this problem, we propose a novel Re-parameterized Feature Distillation Network (ReFDN) for lightweight and efficient RSISR tasks. Feature distillation, refinement, condensation, and enhancement are efficiently integrated into the re-parameterized feature distillation block named ReFDB for lighter and stronger feature extraction. With the help of elaborate re-parameterized convolution (ReConv) design, we further boost the feature refinement capability without extra inference costs. Additionally, we design an efficient channel and spatial attention module (ECSA) to enhance the important objects and regions of the intermediate features adaptively. Conducted on both commonly used datasets and additional Google Earth data, the experimental results demonstrate our method can achieve a good trade-off between SR performance and network complexity. Our code will be publicly available at https://github.com/DaxingZ/ReFDN.

Fine-grained ship classification (FGSCR) has many applications in military and civilian fields. In recent years, deep learning has been widely used for classification tasks, and its success is inseparable from that of big data. However, ship images are valuable, with only a few images of a specific category being obtained, leading to the fine-grained few-shot ship classification problem. In addition, feature map channels contain distinct characteristics and discriminative details, which significantly influence FGSCR. Intuitively, channels with distinct characteristics should be assigned larger weights for classification, but most few-shot learning methods treat the channels equally. Therefore, we propose a generalized ridge-regression-based channelwise feature map weighted reconstruction network to address these issues. First, we reconstruct the query feature map by assigning different weights to the support feature map channels using the generalized ridge regression method. The channels with large discriminative details contribute more toward reconstruction. Second, we propose a support channel weight module to calculate the channel weight matrix used in the generalized ridge regression method. Finally, based on the reconstructed query feature map, we can calculate the reconstruction error. The reconstruction error is adopted as the distance metric. Our proposed method achieves excellent performance on the fine-grained ship, bird, aircraft, and WHU-RS19 datasets compared with other representative few-shot learning methods. Considering the limited studies on the fine-grained few-shot ship classification problem, we believe that our work is of great significance.

Target occlusion is common in satellite videos, which makes object tracking difficult because most state-of-the-art trackers are not robust to occlusion, particularly complete occlusion. In this letter, we propose a novel correlation filter algorithm with an interacting multiple model (IMM) for object tracking in satellite videos that combines the strength of the correlation filter and the IMM. When the target is occluded, we utilize the IMM to predict target position. Therefore, the proposed tracker is robust to occlusion. The experimental results demonstrate that our tracker performs favorably when the target is occluded and achieves excellent performance compared with state-of-the-art methods.

Aircraft classifiers in remotely sensed images based on deep convolutional neural networks play a significant role in military. However, in practical applications, there is a lack of remote sensing fine-grained aircraft data. In this study, we demonstrate that few-shot learning (FSL) can be effectively used for fine-grained identification of aircraft and propose a new classifier-adaptive earth mover’s distance (Adap-EMD) for recognition of few-sample fine-grained aircraft. Adap-EMD consists of an efficient block attention mechanism (EBAM) and an adaptive feature measurement filter (AFMF). The EBAM effectively fuses channel and spatial correlation to capture global features with more pixel-wise relevance and contextual information. The non-parametric AFMF expresses the key information from the adapted emphasizing feature map to achieve a more accurate similarity measurement. Our model outperforms state-of-the-art models on a major few-shot aircraft fine-grained recognition benchmark dataset, introducing only a few additional computations.

The availability of many commercial satellites has created favorable conditions for tracking typical objects in remote sensing sequences, making them widely useful in numerous applications. However, small objects, multiple similar disruptors, background clutter, and occlusion are significant challenges to this field. This study proposes the novel tracker-temporal motion compensation Siamese network (Siam-TMC) for remote sensing tracking. Our method relies on a multidimensional information-aware module and a temporal motion compensation mechanism. Notably, we propose a dual branch-based Dim-Aware module that brings together foreground and high frequency information to distinguish between critical small objects and interferers. In addition, a TMComp mechanism using temporal motion information was designed to mitigate object trajectory drift through the supervision of occlusion detection. Detailed experimental comparisons on a benchmark dataset show that our method outperforms the state-of-the-art tracking models, particularly in occlusion scenarios.

Interpreting airborne remote sensing images plays an important role in aviation control and battlefield situational awareness. However, highly dynamic aircraft detection remains challenging, owing to variable object sizes, flexible attitudes, and motion blur. This study develops a multi-scale airborne aircraft dataset benchmark to overcome aircraft detection challenges, such as high intraclass variance, multiple scales and angles, motion blur, and partial occlusion. We also propose a trained-from-scratch aircraft detector, the local feature correlation single shot multibox detector (LFC-SSD), to detect multi-scale aircraft. The LFC-SSD comprises a local correlation feature extraction module, called “Right-Residual,” and a feature fusion module using a reverse feature pyramid network (R-FPN). Right-Residual extends the global receptive field by aggregating contextual information while learning non-adjacent region features efficiently. R-FPN utilizes multi-path information transfer horizontally with recursive integration to enhance the robust representation of the location information of multi-scale object features. In addition, a specific default boxes method is designed for remote-sensing images of aircraft. Extensive experimental results confirm the significant improvement of the proposed method over several existing state-of-the-art methods.

Fine-grained ship classification plays an important part in many military and civilian applications. However, it is often costly to obtain images of ships, making it difficult to procure large numbers of such images. This difficulty poses challenges to machine learning procedures that require ship images. Commonly, only a few input images are available for certain types of ships, which leads to the poor generalization of trained models. Therefore, few-shot fine-grained ship classification is an important (but significantly challenging) task in machine learning. In this study, we propose a novel foreground-aware feature map reconstruction network (FRN) that is simple, effective, and scalable. We reconstruct the query features from support features using ridge regression and predict the distribution of the categories of query images between the reconstructed and real query features by comparing the weighted distances with foreground weights. The foreground weights indicate the percentages of foreground information in the feature map locations. We propose two methods for calculating the foreground weights: a non-parametric method and a parametric method. Our proposed network achieves state-of-the-art results on both the fine-grained ship classification dataset Fine-Grained Ship Classification in Remote sensing images (FGSCR) and the natural fine-grained bird classification dataset Caltech UCSD Birds (CUB).

Object tracking in satellite videos faces various challenges such as target occlusion, target rotation, and background clutter. This study proposes a correlation particle filter algorithm with motion estimation for object tracking in satellite videos. The tracker, called CPKF, combines the strengths of the correlation, particle, and Kalman filters. Compared with existing tracking methods based on correlation filters, the proposed tracker has three major advantages: (1) Particle sampling, and motion estimation build robustness against partial and complete occlusion. (2) Color histogram model makes it robust to target rotation. (3) Fusion of multiple feature response maps effectively handle background clutter and low contrast. The experimental results demonstrate that the proposed tracking algorithm performs better than state-of-the-art methods.