The YOLO network has been extensively employed in the field of ship detection in optical images. However, the YOLO model rarely considers the global and local relationships in the input image, which limits the final target prediction performance to a certain extent, especially for small ship targets. To address this problem, we propose a novel small ship detection method, which improves the detection accuracy compared with the YOLO-based network architecture and does not increase the amount of computation significantly. Specifically, attention mechanisms in spatial and channel dimensions are proposed to adaptively assign the importance of features in different scales. Moreover, in order to improve the training efficiency and detection accuracy, a new loss function is employed to constrain the detection step, which enables the detector to learn the shape of the ship target more efficiently. The experimental results on a public and high-quality ship dataset indicate that our method realizes state-of-the-art performance in comparison with several widely used advanced approaches.

Compared with anchor-based detectors, anchor-free detectors have the advantage of flexibility and a lower calculation complexity. However, in complex remote sensing scenes, the limited geometric size, weak features of objects, and widely distributed environmental elements similar to the characteristics of objects make small object detection a challenging task. To solve these issues, we propose an anchor-free detector named FE-CenterNet, which can accurately detect small objects such as vehicles in complicated remote sensing scenes. First, we designed a feature enhancement module (FEM) composed of a feature aggregation structure (FAS) and an attention generation structure (AGS). This module contributes to suppressing the interference of false alarms in the scene by mining multiscale contextual information and combining a coordinate attention mechanism, thus improving the perception of small objects. Meanwhile, to meet the high positioning accuracy requirements of small objects, we proposed a new loss function without extra calculation and time cost during the inference process. Finally, to verify the algorithm performance and provide a foundation for subsequent research, we established a dim and small vehicle dataset (DSVD) containing various objects and complex scenes. The experiment results demonstrate that the proposed method performs better than mainstream object detectors. Specifically, the average precision (AP) metric of our method is 7.2% higher than that of the original CenterNet with only a decrease of 1.3 FPS.

Remote sensing target detection technology in cloud and mist scenes is of great significance for applications such as marine safety monitoring and airport traffic management. However, the degradation and loss of features caused by the obstruction of cloud and mist elements still pose a challenging problem for this technology. To enhance object detection performance in adverse weather conditions, we propose a novel target detection method named CM-YOLO that integrates background suppression and semantic context mining, which can achieve accurate detection of targets under different cloud and mist conditions. Specifically, a component-decoupling-based background suppression (CDBS) module is proposed, which extracts cloud and mist components based on characteristic priors and effectively enhances the contrast between the target and the environmental background through a background subtraction strategy. Moreover, a local-global semantic joint mining (LGSJM) module is utilized, which combines convolutional neural networks (CNNs) and hierarchical selective attention to comprehensively mine global and local semantics, achieving target feature enhancement. Finally, the experimental results on multiple public datasets indicate that the proposed method realizes state-of-the-art performance compared to six advanced detectors, with mAP, precision, and recall indicators reaching 85.5%, 89.4%, and 77.9%, respectively.

With the increasing demand for the wide-area refined detection of aircraft targets, remote sensing cameras have adopted an ultra-large area-array detector as a new imaging mode to obtain broad width remote sensing images (RSIs) with higher resolution. However, this imaging technology introduces new special image degradation characteristics, especially the weak target energy and the low signal-to-noise ratio (SNR) of the image, which seriously affect the target detection capability. To address the aforementioned issues, we propose an aircraft detection method for RSIs with low SNR, termed L-SNR-YOLO. In particular, the backbone is built blending a swin-transformer and convolutional neural network (CNN), which obtains multiscale global and local RSI information to enhance the algorithm’s robustness. Moreover, we design an effective feature enhancement (EFE) block integrating the concept of nonlocal means filtering to make the aircraft features significant. In addition, we utilize a novel loss function to optimize the detection accuracy. The experimental results demonstrate that our L-SNR-YOLO achieves better detection performance in RSIs than several existing advanced methods.

In temporal infrared small target detection, it is crucial to leverage the disparities in spatio-temporal characteristics between the target and the background to distinguish the former. However, remote imaging and the relative motion between the detection platform and the background cause significant coupling of spatio-temporal characteristics, making target detection highly challenging. To address these challenges, we propose a network named LMAFormer. First, we introduce a local motion-aware spatio-temporal attention mechanism that aligns and enhances multi-frame features to extract local spatio-temporal salient features of targets while avoiding interference from moving backgrounds. Second, we employ a multi-scale fusion transformer encoder that computes self-attention weights across and within scales during encoding, to establish multi-scale correlations among different regions of temporal images, enabling motion background modeling. Lastly, we propose a multi-frame joint query decoder. The shallowest feature map after multi-scale feature propagation is mapped to initial query weights, which are refined through grouped convolutions to generate grouped query vectors. These are jointly optimized to encapsulate rich multi-frame details, strengthening motion background modeling and target feature representation, improving prediction accuracy. Experimental results on the NUDT-MIRSDT, IRDST and the established TSIRMT datasets demonstrate that our network outperforms state-of-the-art (SOTA) methods. Our code and dataset will be available at https://github.com/lifier/LMAFormer.

Infrared small target detection involves determining the coordinate position of the target in complex infrared images. However, challenges arise due to the absence of internal texture structure, edge dispersion, weak energy characteristics of the target, and a significant amount of background clutter resembling the target’s morphology, impeding precise target location. To address these challenges, we propose an infrared small target detection network, named FDDBA-NET (Frequency Domain Decoupling Bidirectional Interactive Attention Network), designed from the perspective of frequency domain decoupling. To suppress backgrounds that are similar in shape and structure to the target, exploiting the spectral differences between the target and background in the frequency domain, we adopt two learnable masks to extract the target-specific spectrum and the target-background-consistent spectrum detrimental to detection. The specific spectrum aids in target detection. A target-level contrast loss is designed to maximize the disparity between these two spectra, ensuring optimal detection results. Additionally, to preserve target details in high-level semantic information, we introduce a bidirectional interactive attention module that leverages mutual modulation of deep global and shallow local features, facilitating deep and shallow feature fusion. To validate our approach, we conduct experiments comparing our proposed network with state-of-the-art conventional methods and deep learning methods on public datasets. The results demonstrate the superior performance of our method.