The infrared small target detection technology has a wide range of applications in maritime defense warning and maritime border reconnaissance, especially in the maritime and sky scenes for detecting potential terrorist attacks and monitoring maritime borders. However, due to the weak nature of infrared targets and the presence of background interferences such as wave reflections and islands in maritime scenes, targets are easily submerged in the background, making small infrared targets hard to detect. We propose the multidimensional information fusion network(MIFNet) that can learn more information from limited data and achieve more accurate target segmentation. The multidimensional information fusion module calculates semantic information through the attention mechanism and fuses it with detailed information and edge information, enabling the network to achieve more accurate target position detection and avoid detecting one target as multiple ones, especially in high-precision scenes such as maritime target detection, thus effectively improving the accuracy and reliability of detection. Moreover, experiments on our constructed dataset for small infrared targets in maritime scenes demonstrate that our algorithm has advantages over other state-of-the-art algorithms, with an IoU of 79.09%, nIoU of 79.43%, F1 score of 87.88%, and AuC of 95.96%.

Accurate and comprehensive 3D objects detection is important for perception systems in autonomous driving. Nevertheless, contemporary mainstream methods tend to perform more effectively on large objects in regions proximate to the LiDAR, leaving limited exploration of long-range objects and small objects. The divergent point pattern of LiDAR, which results in a reduction in point density as the distance increases, leads to a non-uniform point distribution that is ill-suited to discretized volumetric feature extraction. To address this challenge, we propose the Foreground Voxel Proposal (FVP) module, which effectively locates and generates voxels at the foreground of objects. The outputs are subsequently merged to mitigating the difference in point cloud density and completing the object shape. Furthermore, the susceptibility of small objects to occlusion results in the loss of feature space. To overcome this, we propose the Multi-Scale Feature Integration Network (MsFIN), which captures contextual information at different ranges. Subsequently, the outputs of these features are integrated through a cascade framework based on transformers in order to supplement the object features space. The extensive experimental results demonstrate that our network achieves remarkable results. Remarkably, our approach demonstrated an improvement of 8.56% AP on the SECOND baseline for the Car detection task at a distance of more than 20 m, and 9.38% AP on the Cyclist detection task.

The Wigner-Ville distribution (WVD) is a widely used time-frequency analysis (TFA) technique in non-stationary signal processing. However, as signal bandwidths increase, the low throughput of traditional serial architectures hinders real-time WVD processing. In this paper, we propose a novel dual-channel, fully parallel WVD architecture based on data shifting method. We comprehensively tap into the parallel potential in the WVD algorithm and present the corresponding data flow graph. By utilizing a symmetric dual-channel structure, we significantly streamline the generation and combination of autocorrelation sequences, while halving the total number of complex multiplications and FFTs. Additionally, we develop a memory-based, three-stage autocorrelation generator using a data-shifting technique, thereby obviating the necessity for address generation stages. Implementation results show that our design achieves a 95% latency reduction compared to the state-of-the-art WVD architectures, attaining a maximum clock frequency of 322 MHz on a Xilinx UltraScale+ FPGA. The proposed architecture holds substantial potential for augmenting the real-time processing of wide-band signals.

Infrared small-target tracking is a fundamental yet challenging task in computer vision, owing to extremely small target sizes, severe background clutter, and low signal-to-noise ratios, which often cause tracking drift or complete target loss. As a representative method in this field, SiamDT has achieved promising performance on the challenging Anti-UAV410 benchmark—a large-scale thermal infrared dataset. However, SiamDT still suffers from three key limitations: first, its standard Feature Pyramid Network (FPN) lacks channel-wise selective focusing, causing target-related channels to be overwhelmed by numerous background clutter channels; second, its fixed receptive field cannot adapt to the drastic scale variations in infrared targets; third, its spatial-domain feature fusion lacks high-frequency component compensation, resulting in severe loss of discriminative edge and contour details in deep features. To address these issues, we propose Enhanced-SiamDT, an enhanced tracking model built upon the SiamDT framework. Our contributions are twofold. First, we design a Multi-Attention Feature Pyramid Network (MA-FPN) that sequentially integrates Efficient Channel Attention (ECA), Large Selective Kernel (LSK), and Wavelet Domain Attention (WDAM) to achieve channel-wise recalibration, adaptive multi-scale receptive field selection, and frequency-domain detail enhancement, thereby suppressing background clutter and strengthening small-target feature representations. Second, building upon the dual similarity learning architecture inherited from the SiamDT baseline, we introduce a background prototype suppression strategy that reduces false alarms by penalizing candidate boxes with high similarity to background prototypes and a conservative template update mechanism with explicit update criteria, which prevents template drift under fast motion and short-term occlusion. Extensive experiments on the Anti-UAV410 and Anti-UAV benchmarks demonstrate that Enhanced-SiamDT achieves new state-of-the-art performance, with a State Accuracy (SA) of 68.58% and 71.84%, and a Precision of 89.92% and 92.71%, respectively. These results validate that our approach effectively overcomes the limitations of existing Siamese trackers, delivering significant improvements in discriminative feature extraction and tracking robustness for infrared small targets.