Abstract:Synthetic aperture radar (SAR) images obtained from multi-sensor systems usually exhibit significant shift in data distribution, known as the domain shift. It is challenging to utilize the relevant knowledge of multi-sensor datasets for SAR image cross-domain classification with general supervised learning methods. While domain adaptation (DA) methods can alleviate the domain shift by aligning distributions, they are limited in handling the scale/resolution variations observed in multi-sensor SAR images. These methods mainly focus on feature representations at a single scale for distribution alignment, which may fail to fully align the distributions across different scales. To solve this problem, we propose a new multi-scale alignment domain adaptation network (MSADAN) for SAR ship cross-domain classification. MSADAN explicitly considers the scale factor, enabling us to overcome the weak generalization observed in existing DA methods when dealing with SAR images exhibiting significant scale/resolution variations. Specifically, we develop a scale-aware feature extractor to effectively capture the multi-scale information present in the datasets, facilitating comprehensive representation learning. Furthermore, we propose a multi-level bilinear fusion (MLBF)-based adversarial learning strategy to overcome the limitations of single-scale feature extraction for distribution alignment, aiming to enhance the generalization ability of the model across domains. In addition, a customized class contrastive loss is designed to improve the inter-class separability and intra-class compactness by penalizing cross-class confusion. Experimental results on datasets demonstrate the superiority of MSADAN in SAR ship cross-domain classification. Note to Practitioners —Ship classification using Synthetic aperture radar (SAR) images proves to be a promising strategy in modern maritime monitoring systems. The primary motivation of this paper is to develop a cross-domain classification system tailored for SAR ship target. In real-world scenarios, SAR data often presents significant resolution variations due to the different imaging modes and conditions across multiple sources. Existing DA methods fail to effectively address this unique challenge in the remote sensing field, resulting in poor cross-domain classification performance. The proposed method considers the scale or resolution variations of the targets during feature learning. Furthermore, explicitly incorporating resolution into the distribution alignment enhances the generalization ability of the model to target tasks when dealing with the significant scale or resolution variations across datasets. Experimental results confirm the practicality and robustness of our SAR ship cross-domain classification method in real scenarios. This is of significant importance in promoting the application of machine learning approaches in practical scenarios. In the future, we plan to integrate the physical scattering information of SAR images for the interpretable cross-domain classification for SAR targets.

Adversarial Deep Domain Adaptation for Multi-Band SAR Images Classification

Unsupervised Domain Adaptation for SAR Target Detection

Simulation-assisted SAR Target Classification Based on Unsupervised Domain Adaptation and Model Interpretability Analysis

Multiscale Attention-Based Subdomain Dynamic Adaptation for Cross-Domain Scene Classification

Unsupervised Aircraft Detection in SAR Images with Image-Level Domain Adaption from Optical Images

Unsupervised Domain Adaptation for SAR Target Classification Based on Domain- and Class-Level Alignment: from Simulated to Real Data

Active Learning SAR Image Classification Method Crossing Different Imaging Platforms

Heterogeneous Image Classification with Multi-Stage Conditional Adversarial Domain Adaptation Between SAR and Optical Imagery

Pixel-Level and Feature-Level Domain Adaptation for Heterogeneous SAR Target Recognition.

Selecting Pseudo Supervision for Unsupervised Domain Adaptive SAR Target Classification

A domain adaptation network for cross-imaging satellites sar image ship classification

Multi-scale Deep Feature Learning Network with Bilateral Filtering for SAR Image Classification

Multi-Step Unsupervised Domain Adaptation in Image and Feature Space for Synthetic Aperture Radar Image Terrain Classification

Cross Sensor Transfer Learning for Unsupervised SAR Target Detection

Multi-Scale Alignment Domain Adaptation for Ship Classification in Multi-Resolution SAR Images

Feature Aggregation and Compensation-Based Domain Adaptation Framework for Cross-Resolution Target Recognition in SAR Imagery

A CA-Based Weighted Clustering Adversarial Network for Unsupervised Domain Adaptation PolSAR Image Classification

Unsupervised Cross-Domain Network Based on Bi-classifier for PolSAR Image Classification

An Automatic Ship Detection Method Adapting to Different Satellites SAR Images With Feature Alignment and Compensation Loss

VSFA: Visual and Scattering Topological Feature Fusion and Alignment Network for Unsupervised Domain Adaptation in SAR Target Recognition

A Universal Adversarial Attack on CNN-SAR Image Classification by Feature Dictionary Modeling.