Detecting ships from satellite imagery is vital for maritime surveillance. Most current methods rely on deep learning (DL), which requires a large number of high-quality annotated images to train accurate models. Since satellite imagery comes from various sensors, DL-based ship detection algorithms need to perform well across different sensor types. However, privacy concerns, especially with commercial images, limit data, and annotation sharing. Federated learning (FL) offers a promising solution for collaborative learning while addressing these concerns. Despite its potential, research on FL for ship detection is still sparse. This study implements and evaluates three FL models for detecting ships using multi-source optical satellite images, spanning high to low resolution. Our experiments on two distinct datasets demonstrate that FL models significantly enhance detection performance without centralizing data. Source codes are publicly available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/ffyyytt/FLYOLO</uri>.

The use of scattering models with dielectric surfaces in hybrid-polarimetry (hybrid-pol) synthetic aperture radar (SAR) decomposition methods offers a more realistic representation of land-cover surfaces. Additionally, redefining an appropriate volume scattering model improves classification accuracy, particularly in complex urban areas where correctly classifying pixels in the double-bounce (DB) scattering mechanism category is challenging. In this letter, a prominent volume scattering model is incorporated into the hybrid-pol three-component scattering model (HTM) framework. In this investigation, the proposed algorithm is referred to as “Modified HTM.” The proposed method is based on the Neumann volume scattering model (NVSM). Given the complex polarimetric radar responses from vegetation, NVSM has been demonstrated to be an optimal parameter inversion framework that is capable of efficiently recover the spatial distribution and scattering properties from the volumetric media. To present the improved efficacy, the proposed technique is compared to the conventional hybrid-pol techniques, specifically, the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$m-\delta $ </tex-math></inline-formula>, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$m-\alpha $ </tex-math></inline-formula>, and original HTM decomposition. An assessment based on ship detection has also been included to illustrate its efficacy.

Deblurring in remote sensing images is a challenging task due to the long-range imaging capabilities of remote sensing sensors, which often results in image blur. Factors contributing to image blur include atmospheric disturbances during long-range imaging or the orbital motion of remote sensing platforms. The existing methods remove blur in remote sensing images using the traditional attention mechanism, which focuses on a limited number of features. However, they often overlook the features among neighboring positions in blurry areas, and these areas contain more relevant features. Leveraging these features can effectively assist in restoring the complex object textures of remote sensing blurry images. To achieve this, we propose a novel implicit neural attention mechanism for assembling more relevant features implied by surrounding coordinates. Specifically, we use the features and their corresponding coordinates to learn the enhanced feature representation with more relevant features, and this representation can be used to derive the deblurred images. Extensive experiments demonstrate that our proposed method, INA-RSDeblur, outperforms the state-of-the-art deblurring methods in remote sensing blurry images.

This work introduces the MUlti-modal MUlti-class Change Detection (MUMUCD) dataset, which comprises 70 globally distributed georeferenced bitemporal pairs obtained by multiple space-borne sensors. Acquisitions over heterogenous terrains (e.g., urban, rural, forests, deserts) for the period 2019-2024 are processed to create the first large scale curated dataset combining Synthetic Aperture Radar (Sentinel-1), multispectral (Sentinel-2) and hyperspectral (PRISMA) data with ancillary information. Provided with a resampled image resolution of 10 m and a size of 1536×1536 square pixels, MUMUCD allows to extract several thousands non-overlapping patches with size 128×128 square pixels, enabling data-intensive machine learning (ML) applications. Seasonality plays a critical role in the analysis of environmental data, so we carefully selected scenes representing all times of the year and a variety of geographical contexts relevant to key impact sectors, taking also into account the availability of PRISMA acquisitions. While scenes with low cloudiness are prioritized, cloudy pixels are not excluded as different data combinations (e.g., SAR/optical) can be exploited to mitigate the atmospheric effects. Beyond coregistered data, we include surface elevation, land cover and binary change maps obtained by processing the Dynamic World dataset, based on the provided multispectral data. Benchmarking of machine and deep learning algorithms indicates that the provided labels should be augmented to get the most out of the multiple modalities. MUMUCD is meant to fill a research gap by offering multi-sensor and task-oriented data, which make it ideal to fine-tune standard and foundational ML models for a variety of tasks, and even unique when these tasks involve change detection or hyperspectral measurements.

The traditional 2-D multiplication modulation jamming (2D-MMJ) method is widely used in synthetic aperture radar (SAR) deception jamming due to its simple hardware implementation and convenient template preparation. However, it suffers from serious defocusing in high-resolution spaceborne SAR deception jamming. To improve the fidelity of deception jamming, this letter proposes an improved 2D-MMJ method based on defocus correction (2D-MMJDC). First, the defocus correction phase is designed by analyzing the range cell migration (RCM) curves between the real and jamming signals. Next, considering the false point-target azimuth deviation caused by the defocus correction, the azimuth deviation correction phase is designed. Finally, two phases are added to the 2D-MMJ method, and a template preparation process based on the 2-D fast Fourier transform (2D-FFT) is designed. Simulation results show that the 2D-MMJDC method effectively solves the problem of defocusing on the 2D-MMJ method and greatly improves the fidelity of the deception jamming.

Although Mamba models significantly improve hyperspectral image (HSI) classification, one critical challenge is the difficulty in building the sequence of Mamba tokens efficiently. This letter presents a Sparse Deformable Mamba (SDMamba) approach for enhanced HSI classification, with the following contributions. First, to enhance Mamba sequence, an efficient Sparse Deformable Sequencing (SDS) approach is designed to adaptively learn the ”optimal" sequence, leading to sparse and deformable Mamba sequence with increased detail preservation and decreased computations. Second, to boost spatial-spectral feature learning, based on SDS, a Sparse Deformable Spatial Mamba Module (SDSpaM) and a Sparse Deformable Spectral Mamba Module (SDSpeM) are designed for tailored modeling of the spatial information spectral information. Last, to improve the fusion of SDSpaM and SDSpeM, an attention based feature fusion approach is designed to integrate the outputs of the SDSpaM and SDSpeM. The proposed method is tested on three benchmark datasets with many state-of-the-art approaches, inlcuding CNNs, GAN Transformer, and Mamba based methods, demonstrating that the proposed approach can achieve higher accuracy with less computation, and better detail small-class preservation capability.

The global demand for unconventional energy sources such as geothermal energy and white hydrogen requires new exploration techniques for precise subsurface structure characterization and potential reservoir identification. The magnetotelluric (MT) method is crucial for these tasks, providing critical information on the distribution of subsurface electrical resistivity at depths ranging from hundreds to thousands of meters. However, traditional iterative algorithm-based inversion methods require the adjustment of multiple parameters, demanding time-consuming and exhaustive tuning processes to achieve proper cost function minimization. Recent advances have incorporated deep learning algorithms for MT inversion, primarily based on supervised learning, and large labeled datasets are needed for training. This work utilizes TensorFlow operations to create a differentiable forward MT operator, leveraging its automatic differentiation capability. Moreover, instead of solving for the subsurface model directly, as classical algorithms perform, this letter presents a new deep unsupervised inversion algorithm guided by physics to estimate 1-D MT models. Instead of using datasets with the observed data and their respective model as labels during training, our method employs a differentiable modeling operator that physically guides the cost function minimization, making the proposed method solely dependent on observed data. Therefore, the optimization algorithm updates the network weights to minimize the data misfit. We test the proposed method with field and synthetic data at different acquisition frequencies, demonstrating that the resistivity models obtained are more accurate than those calculated using other techniques. Our implementation is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/PAULGOYES/MT_guided1DInversion.git</uri>.

For large-scale geophysical models, the order of the coefficient matrix in 3-D transient electromagnetics (TEMs) forward modeling can reach millions or even tens of millions. Balancing computational efficiency and memory usage presents a challenge worthy of in-depth exploration. In this letter, we utilize an integral representation of the iterative error in the Arnoldi method to construct an efficient quadrature-based restarted forward algorithm. First, the mimetic finite volume (MFV) method on a staggered hexahedral grid is employed to discretize the time-domain Maxwell’s equations, expressing the TEM response after the step-off waveform shutoff as the product of the matrix exponential function <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$f({\text {A}})$ </tex-math></inline-formula> and vector b. Then, using Cauchy’s integral formula, the expression of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${f}({\text {A}}){b}$ </tex-math></inline-formula> is transformed into an integral form and approximated using the restarted Arnoldi (RA) algorithm. Our method does not require solving linear systems and can leverage GPU parallel technology and optimize the RA algorithm parameters to enhance computational efficiency. Comparative studies with other numerical methods validate the advantages and accuracy of our approach, which numerical example demonstrates can fully achieve large-scale fast 3-D TEM forward modeling.

Full-waveform airborne LiDAR bathymetry (ALB), which provides waveforms and point clouds, has become an essential technology for shallow water surveys. However, weak seafloor echoes are challenging to detect accurately because of waveform feature confusion caused by the complex measurement environments. To address this issue, waveform feature importances, feature histograms, and feature spaces of 14-dimensional waveform features are conducted to analyze the waveform feature confusion. Then, a random forest with optimized thresholds (RFOT) is proposed to detect normal seafloor echoes and weak seafloor echoes. Finally, waveform sharpening and condition screening are used to extract the seafloor echoes for overlapping waveforms in very shallow waters. The proposed method was verified with 14 swaths obtained by the Optech Aquarius system around Wuzhizhou Island. The results show that the energy features (area under curve, amplitude, etc.) can better discriminate the difference between weak seafloor echoes and noise than the shape features (RL area ratio, kurtosis, etc.). The number of seafloor echoes detected by the proposed method increased by 148.86% compared with the Aquarius system. The reference data prove that seafloor points detected by the proposed method are accurate and effective. Thus, this contribution effectively improves the bathymetric performance of the ALB system.

Visual tracking on unmanned aerial vehicle (UAV) platforms is a fundamental and crucial visual task. Compared to conventional tracking tasks, aerial tracking faces specific challenging scenarios due to its unique perspective. Although existing aerial trackers have demonstrated promising performance, they remain limited in capturing spatial-channel saliency across branches and effectively utilizing historical information. To address these issues, this letter proposes a spatial-channel saliency and temporal fusion network (SiamSCT), which aims to enhance feature representation for efficient and accurate aerial tracking. Specifically, SiamSCT introduces a weight-shared spatial saliency block to strengthen the spatial feature representation across the tracking network’s branches. In addition, a light channel aware module is designed to facilitate deep feature interaction across branches at the channel level, further enhancing feature discriminability. Finally, using a historical similarity response fusion strategy, SiamSCT achieves more stable and reliable tracking responses, effectively tackling complex aerial scenarios. Extensive experiments on several authoritative aerial tracking datasets demonstrate that SiamSCT outperforms state-of-the-art trackers. Furthermore, SiamSCT achieves a tracking speed of 133 FPS on NVIDIA RTX 3060Ti, proving its excellent performance in real-time.