Generating pixel-level annotations for semantic segmentation tasks of high-resolution remote sensing images is both time-consuming and labor-intensive, which has led to increased interest in unsupervised methods. Therefore, in this paper, we propose an unsupervised multi-scale hybrid feature extraction network based on the CNN-Transformer architecture, referred to as MSHFE-Net. The MSHFE-Net consists of three main modules: a Multi-Scale Pixel-Guided CNN Encoder, a Multi-Scale Aggregation Transformer Encoder, and a Parallel Attention Fusion Module. The Multi-Scale Pixel-Guided CNN Encoder is designed for multi-scale, fine-grained feature extraction in unsupervised tasks, efficiently recovering local spatial information in images. Meanwhile, the Multi-Scale Aggregation Transformer Encoder introduces a multi-scale aggregation module, which further enhances the unsupervised acquisition of multi-scale contextual information, obtaining global features with stronger feature representation. The Parallel Attention Fusion Module employs an attention mechanism to fuse global and local features in both channel and spatial dimensions in parallel, enriching the semantic relations extracted during unsupervised training and improving the performance of unsupervised semantic segmentation. K-means clustering is then performed on the fused features to achieve high-precision unsupervised semantic segmentation. Experiments with MSHFE-Net on the Potsdam and Vaihingen datasets demonstrate its effectiveness in significantly improving the accuracy of unsupervised semantic segmentation.

Complex-valued convolutional neural networks (CVCNNs) have demonstrated strong capabilities for polarimetric synthetic aperture radar (PolSAR) image classification by effectively integrating both amplitude and phase information inherent in polarimetric data. However, their practical deployment faces significant challenges due to high computational costs and performance degradation caused by extremely limited labeled samples. To address these challenges, a lightweight CV Siamese network (LCVSNet) is proposed for few-shot PolSAR image classification. Considering the constraints of limited hardware resources in practical applications, simple one-dimensional (1D) CV convolutions along the scattering dimension are combined with two-dimensional (2D) lightweight CV convolutions. In this way, the inter-element dependencies of polarimetric coherency matrix and the spatial correlations between neighboring units can be captured effectively, while simultaneously reducing computational costs. Furthermore, LCVSNet incorporates a contrastive learning (CL) projection head to explicitly optimize the feature space. This optimization can effectively enhance the feature discriminability, leading to accurate classification with a limited number of labeled samples. Experiments on three real PolSAR datasets demonstrate the effectiveness and practical utility of LCVSNet for PolSAR image classification with a small number of labeled samples.

The powerful graph convolutional network (GCN) for polarimetric synthetic aperture radar (PolSAR) image classification generally relies on real-valued features, ignoring the phase information and thus limiting the modeling of complex-valued (CV) polarization characteristics. To address this issue, this paper proposes a novel multiscale attention-enhanced CV graph U-Net model, abbreviated as MAE-CV-GUNet, by embedding CV-GCN into a graph U-Net framework augmented with multiscale attention mechanisms. First, a CV-GCN is constructed based on the real-valued GCN, to effectively capture the intrinsic amplitude and phase information of the PolSAR data, along with the underlying correlations between them. This way can well lead to an improved feature representation for PolSAR images. Based on CV-GCN, a CV graph U-Net (CV-GUNet) architecture is constructed by integrating multiple CV-GCN components, aiming to extract multi-scale features and further enhance the ability to extract discriminative features in the complex domain. Then, a multiscale attention (MSA) mechanism is designed, enabling the proposed MAE-CV-GUNet to adaptively learn the importances of features at various scales, thereby dynamically fusing the multiscale information among them. The comparisons and ablation experiments on three PolSAR datasets show that MAE-CV-GUNet has excellent performance in PolSAR image classification.