Automatic ship detection in SAR images plays an essential role in both military and civilian fields. However, most of the existing deep learning detection methods introduce complex models and huge calculations while improving the detection accuracy, which is not conducive to the application of real-time ship detection. To solve this problem, an efficient lightweight network YOLO-Lite is proposed for SAR ship detection in this paper. First, a lightweight feature enhancement backbone (LFEBNet) is designed to reduce the amount of calculation. Additionally, a channel and position enhancement attention (CPEA) module is constructed and embedded into the backbone network to more accurately locate the target location by capturing the positional information. Second, an enhanced spatial pyramid pooling (EnSPP) module is customized to enhance the expression ability of features and address the position information loss of small SAR ships in high-level features. Third, we construct an effective multi-scale feature fusion network (MFFNet) with two feature fusion channels to obtain feature maps with more position and semantic information. Furthermore, a novel confidence loss function is proposed to effectively improve the SAR ship target detection accuracy. Extensive experiments on SSDD and SAR ship datasets verify the effectiveness of our YOLO-Lite, which can not only accurately detect SAR ships in different backgrounds but can also realize a lightweight architecture with low computation cost.

It is necessary to detect the target reflections in ground penetrating radar (GPR) images, so that surface metal targets can be identified successfully. In order to accurately locate buried metal objects, a novel method called the Multiresolution Monogenic Signal Analysis (MMSA) system is applied in ground penetrating radar (GPR) images. This process includes four steps. First the image is decomposed by the MMSA to extract the amplitude component of the B-scan image. The amplitude component enhances the target reflection and suppresses the direct wave and reflective wave to a large extent. Then we use the region of interest extraction method to locate the genuine target reflections from spurious reflections by calculating the normalized variance of the amplitude component. To find the apexes of the targets, a Hough transform is used in the restricted area. Finally, we estimate the horizontal and vertical position of the target. In terms of buried object detection, the proposed system exhibits promising performance, as shown in the experimental results.

Down-looking MIMO array SAR can reconstruct 3D images of the observed area in the inferior of the platform of the SAR and has wide application prospects. In this paper, a new strategy based on Bayesian compressive sensing theory is proposed for down-looking MIMO array SAR imaging, which transforms the cross-track imaging process of down-looking MIMO array SAR into the problem of sparse signal reconstruction from noisy measurements. Due to account for additive noise encountered in the measurement process, high quality image can be achieved. Simulation results indicate that the proposed method can provide better resolution and lower sidelobes compared to the conventional method.

SAR ship detection plays a crucial role in various fields, such as maritime transportation supervision and crisis rescue. However, in most SAR ship images, the targets occupy a small number of pixels in the image, making it difficult to distinguish the targets from the background. This challenge makes it hard for lightweight deep learning-based object detection algorithms to obtain effective feature information. To address these issues, a SAR ship detection network with effective feature representation is proposed. Initially, an efficient feature extraction network (EFENet) using a hybrid architecture of convolutional neural network and Vision Transformer is proposed. The EFENet learns both local and global information within images, while embedding a bidirectional attention mechanism (BAM) module for precise target localization and enhanced capture of small target information. Subsequently, the multi-branch spatial pyramid pooling module (MBSPP) is introduced to expand the receptive field of feature maps, addressing the issue of the target position information loss in high-level features. Lastly, a cross-scale feature fusion network (CSFNet) is designed to aggregate feature maps of different scales, resulting in feature maps containing rich semantic and positional information. Experimental results demonstrate the effectiveness of this approach in achieving high-quality detection results on both the SSDD dataset and the HRSID dataset.

Recent studies have demonstrated that machine learning-based generative models can create novel molecules with desirable properties. Among them, Conditional Variational Autoencoder (CVAE) is a powerful approach to generate molecules with desired physiochemical and pharmacological properties. However, the CVAE’s latent space is still a black-box, making it difficult to understand the relationship between the latent space and molecular properties. To address this issue, we propose the Interpretable Conditional Variational Autoencoder (ICVAE), which introduces a modified loss function that correlates the latent value with molecular properties. ICVAE established a linear mapping between latent variables and molecular properties. This linearity is not only crucial for improving interpretability, by assigning clear semantic meaning to latent dimensions, but also provides a practical advantage. It enables direct manipulation of molecular attributes through simple coordinate shifts in latent space, rather than relying on opaque, black-box optimization algorithms. Our experimental results show that the ICVAE can linearly relate one or multiple molecular properties with the latent value and generate molecules with precise properties by controlling the latent values. The ICVAE’s interpretability allows us to gain insight into the molecular generation process, making it a promising approach in drug discovery and material design.

Terahertz time domain spectroscopy imaging systems suffer from the problems of long image acquisition time and massive data processing. Reducing the sampling rate will lead to the degradation of the imaging reconstruction quality. To solve this issue, a novel terahertz imaging model, named the dual sparsity constraints terahertz image reconstruction model (DSC-THz), is proposed in this paper. DSC-THz fuses the sparsity constraints of the terahertz image in wavelet and gradient domains into the terahertz image reconstruction model. Differing from the conventional wavelet transform, we introduce a non-linear exponentiation transform into the shift invariant wavelet coefficients, which can amplify the significant coefficients and suppress the small ones. Simultaneously, the sparsity of the terahertz image in gradient domain is used to enhance the sparsity of the image, which has the advantage of edge preserving property. The split Bregman iteration scheme is utilized to tackle the optimization problem. By using the idea of separation of variables, the optimization problem is decomposed into subproblems to solve. Compared with the conventional single sparsity constraint terahertz image reconstruction model, the experiments verified that the proposed approach can achieve higher terahertz image reconstruction quality at low sampling rates.