In recent years, remote sensing (RS) vision foundation models such as RingMo have emerged and achieved excellent performance in various downstream tasks. However, the high demand for computing resources limits the application of these models on edge devices. It is necessary to design a more lightweight foundation model to support on-orbit RS image interpretation. Existing methods face challenges in achieving lightweight solutions while retaining generalization in RS image interpretation. This is due to the complex high and low-frequency spectral components in RS images, which make traditional single CNN or Vision Transformer methods unsuitable for the task. Therefore, this paper proposes RingMo-lite, a RS lightweight network with a CNN-Transformer hybrid framework, which effectively exploits the frequency-domain properties of RS to optimize the interpretation process on several tasks like classification, object detection, semantic segmentation, and change detection. It is combined by the Transformer module as a low-pass filter to extract global features of RS images through a dual-branch structure, and the CNN module as a stacked high-pass filter to extract fine-grained details effectively. Furthermore, a novelty-designed frequency-domain masked image modeling (FD-MIM) is employed during the pretraining stage for self-supervised learning, which combines the high-frequency and low-frequency characteristics of each image patch. This approach effectively captures the latent feature representation in RS data. As shown in Fig. 1, compared with RingMo, the proposed RingMo-lite reduces the parameters over 60% in various RS image interpretation tasks, the average accuracy drops by less than 2% in most of the scenes and achieves SOTA performance compared to models of the similar size. In addition, our work will be integrated into the MindSpore computing platform in the near future.

With the development of information technology, multi-platform collaborative collection and processing of remote sensing images has become a significant trend. However, the existing models are challenging to achieve accurate and efficient image interpretation on remote sensing multi-platform systems. To solve this problem, we propose a novel distributed convolutional neural network (DCNNet) and demonstrate the superiority of our method in remote sensing image classification. Firstly, a progressive inference mechanism is introduced to support most images to be classified in advance with satisfactory accuracy, which minimises redundant cloud transmission and achieves higher inference acceleration. Meanwhile, a distributed self-distillation paradigm is designed to integrate and refine in-depth features, performing efficient knowledge transfer between terminals and the cloud network. Secondly, a multi-scale feature fusion (MSFF) module is presented to extract valid receptive fields and assign weights to crucial channel dimension features. Finally, a sampling augmentation (SA) attention is proposed to enhance the effective feature representation of RS images through a bottom-up and top-down feedforward structure. We conducted extensive experiments and visual analyses on three benchmark scene classification datasets and one fine-grained dataset. Compared with the existing methods, DCNNet consolidates several advantages in terms of accuracy, computation, transmission and processing efficiency into a single framework for multi-platform remote sensing image classification.

Multi-UAV collaborative 3D object detection can comprehend complex environments by integrating complementary information, with applications encompassing traffic monitoring, delivery services and agricultural management. However, the extremely broad observations in aerial remote sensing and significant perspective differences across multiple UAVs make it challenging to achieve precise and consistent feature mapping from 2D images to 3D space in multi-UAV collaborative 3D object detection paradigm. To address the problem, we propose an unparalleled camera-based multi-UAV collaborative 3D object detection paradigm called UCDNet. Specifically, the depth information from the UAVs to the ground is explicitly utilized as a strong prior to provide a reference for more accurate and generalizable feature mapping. Additionally, we design a homologous points geometric consistency loss as an auxiliary self-supervision, which directly influences the feature mapping module, thereby strengthening the global consistency of multi-view perception. Experiments on AeroCollab3D and CoPerception-UAVs datasets show our method increases 4.7% and 10% mAP respectively compared to the baseline, which demonstrates the superiority of UCDNet.

Large-size images cannot be directly put into GPU for training and need to be cropped to patches due to GPU memory limitation. The commonly used cropping methods before are random cropping and sequential cropping, which are crude and fatally inefficient. Firstly, categories of datasets are often imbalanced, and just simple cropping misses an excellent opportunity to make the data distribution balanced. Secondly, the training needs to crop a large number of patches to cover all patterns, which greatly increases the training time. This problem is of great practical hazards but is often overlooked by previous works. The optimal solution is to generate valuable patches. Valuable patches refer to the value to network training, i.e., the value of this patch for the convergence of the network, and the improvement of the accuracy. To this end, we propose a data-related patch proposal strategy to sample high valuable patches. The core idea is to score each patch according to the accuracy of each category, so as to perform balanced sampling. Compared with random cropping or sequential cropping, our method can improve the segmentation accuracy and accelerate the training vastly. Moreover, our method also shows great advantages over the loss-based balanced approaches. Experiments on Deepglobe and Potsdam show the excellent effect of our method.