Hyperspectral image classification (HSIC) is one of the most important research topics in the field of remote sensing. However, it is difficult to label hyperspectral data, which limits the improvement of classification performance of hyperspectral images in the case of small samples. To alleviate this problem, in this paper, a dual-branch network which combines cross-channel dense connection and multi-scale dual aggregated attention (CDC_MDAA) is proposed. On the spatial branch, a cross-channel dense connections (CDC) module is designed. The CDC can effectively combine cross-channel convolution with dense connections to extract the deep spatial features of HSIs. Then, a spatial multi-scale dual aggregated attention module (SPA_MDAA) is constructed. The SPA_MDAA adopts dual autocorrelation for attention modeling to strengthen the differences between features and enhance the ability to pay attention to important features. On the spectral branch, a spectral multi-scale dual aggregated attention module (SPE_MDAA) is designed to capture important spectral features. Finally, the spatial spectral features are fused, and the classification results are obtained. The experimental results show that the classification performance of the proposed method is superior to some state-of-the-art methods in small samples and has good generalization.

Remote sensing images obtained at high altitudes often contain complete object or scene information, which makes their global visual features richer compared to natural images. In order to enhance the scope and multilevel characteristics of global visual features of remote sensing images, this article proposes an enhanced global feature-guided network based on multiple filtering noise reduction (GFRNet) for remote sensing image compression. First, a pyramid vision transformer (PVT) is introduced into remote sensing image compression for the first time. Based on this, a PVT compression branch (PVTCB) is designed, which can capture multilevel global visual features through a three-stage pyramid transformer module for image compression (TPTC) and utilizes filters to accurately control the output of TPTC. Second, a quadruple-filtered multicore noise reduction attention module (QFMR-AM) is constructed in the four-stage compression branch (FSCB) for denoising and enhancing multilevel features. Finally, a global visual feature guidance module (GVGM) is designed between FSCB and the four-stage reconstruction decoder (FSRD). By calculating the global visual feature loss LossGVF through GVGM, a novel rate-distortion LossTotal is constructed, making the network more focused on extracting global information. Experimental results show that compared with some advanced methods, the proposed GFRNet achieves better compression performance on multiple evaluation indicators. In addition, the reconstructed images obtained by the proposed GFRNet can provide better classification performance, which further proves that the proposed method helps to preserve more important features of remote sensing images during the compression process.

Infrared Small Target Detection (IRSTD) holds significant application value in military operations, early warning and surveillance, aerospace, and other fields. However, traditional detection methods face challenges related to small target pixel sizes, strong background noise, and sparse infrared image features, resulting in insufficient accuracy and robustness. This paper proposes IR-WSANet, a lightweight network based on an improved YOLOv10, which enhances the detection performance of infrared small targets through a frequency-spatial joint optimization strategy. Firstly, discrete wavelet transform convolution (DWaveletConv) is introduced into the backbone network, and the fusion of high-frequency details and low-frequency semantics is enhanced by multi-band feature decomposition to suppress noise interference; Secondly, we designed a cooperative module (POS-SHSA) that integrates POSConvEmbedding with a partial channel single-head self-attention mechanism (SHSA), which combines local spatial features and global context information to improve the positioning accuracy of small targets. Experiments verify the effectiveness of the model on SIDD and HIT-UAV datasets: the mAP of IR-WSANet on SIDD-City, SIDD-Mountain and HIT-UAV datasets reaches 97.2%, 82.6% and 82.8%, respectively, which is 2.8% to 14.1% higher than the baseline YOLOv10, and the highest F1 score was improved to 14.8%, while maintaining low computing cost (27.9 GFLOPs) and real-time performance (42.8 FPS). The results show that IR-WSANet significantly improves the detection performance of infrared small targets in complex scenes through the combination of frequency domain filtering enhancement and space-channel dual attention mechanism.

Infrared Small Target Detection (IRSTD) holds significant application value in military operations, early warning and surveillance, aerospace, and other fields. However, traditional detection methods face challenges related to small target pixel sizes, strong background noise, and sparse infrared image features, resulting in insufficient accuracy and robustness. This paper proposes IR-WSANet, a lightweight network based on an improved YOLOv10, which enhances the detection performance of infrared small targets through a frequency-spatial joint optimization strategy. Firstly, the discrete wavelet transform convolution (DWaveletConv) is introduced into the backbone network, and the fusion of high-frequency details and low-frequency semantics is enhanced by multi-band feature decomposition to suppress noise interference. Secondly, we designed a cooperative module (POS-SHSA) that integrates POSConvEmbedding with a partial channel single-head self-attention mechanism (SHSA), which combines local spatial features and global context information to improve the positioning accuracy of small targets. Experiments verify the effectiveness of the model on SIDD and HIT-UAV datasets: the mAP of IR-WSANet on SIDD-City, SIDD-Mountain and HIT-UAV datasets reaches 97.2%, 82.6% and 82.8%, respectively, which is 2.8% to 14.1% higher than the baseline YOLOv10, and the highest F1 score was improved to 14.8%, while maintaining low computing cost (27.9 GFLOPs) and real-time performance (42.8 FPS). The results show that IR-WSANet significantly improves the detection performance of infrared small targets in complex scenes through the combination of frequency domain filtering enhancement and space-channel dual attention mechanism.

In this paper, a low-profile broadband and high-gain circularly polarized (CP) metasurface antenna (MSA) is proposed. The characteristic mode analysis is employed to select the valuable modes for the 4×4 square metasurface elements printed on the top of the single-layer dielectric slab. The coplanar waveguide feeding structure combines a horizontal aperture and two slots rotated in a counter-clockwise direction is utilized to excite the CP radiation of the MSA. The proposed CP MSA is fabricated and measured, which achieves a -10 dB impedance bandwidth ranging from 4.83 GHz to 6.35 GHz, with a fractional bandwidth of 27.2%, the overlapped 3 dB axial ratio (AR) bandwidth is 20.1% (covers from 4.83 GHz to 5.91 GHz). Furthermore, the peak boresight gain measured at 5.2 GHz reaches 9.88 dBic. The average gain consistently maintains 8.94 dBic throughout the overlapped AR bandwidth, while the 3 dB gain bandwidth fully encompasses the entirety of the 3 dB overlapped AR bandwidth showing good performance for 5G Wi-Fi band utilization.