Abstract:Supervised deep neural networks (DNNs) have proven to be powerful tools for synthetic aperture radar (SAR) image interpretation tasks. However, they present a formidable challenge in acquiring a substantial amount of labeled data. In this article, we investigate the promising technique of contrastive self-supervised learning for SAR image classification. This approach allows us to take advantage of a large number of available unlabeled images to pretrain an SAR image classification model. Our novel contrastive learning framework conducts both instance-level and cluster-level pretext tasks, which not only enforce consistency between the images and their augmented “views” at the instance level but also their representation within clusters. Besides generating different views through random, low-level image transformations, we proposed two new strategies to construct positive sample pairs to improve contrastive SAR image feature learning: middle-level optical assistance and high-level graph searching. The middle-level optical assistance strategy is inspired by the observation that domain experts typically interpret SAR images with the aid of optical images. This insight spurs us to generate intermediate SAR images as positive samples from geographically matched optical data using CycleGAN. Furthermore, we augment the positive samples of the image with their $k$ -nearest neighbor (KNN) counterparts, following the idea that the KNN samples should belong to the same cluster. Extensive experimental results conducted on the SEN12MS land cover classification benchmark dataset demonstrate that our method is competitive with state-of-the-art self-supervised methods for SAR image classification. Even with only a small amount of labeled data for fine-tuning the model, our method rapidly improves classification performance, surpassing models pretrained on natural image datasets.

SAR Image Classification Based on the Multi-Layer Network and Transfer Learning of Mid-Level Representations.

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

The Classification of Synthetic Aperture Radar Image Target Based on Deep Learning.

SAR Image Classification Via Deep Recurrent Encoding Neural Networks

Synthetic aperture radar remote sensing scene classification based on fuzzy co-occurrence networks

High-Resolution SAR Image Classification Via Deep Convolutional Autoencoders

Classification of Large-Scale High-Resolution SAR Images with Deep Transfer Learning

Deep Supervised and Contractive Neural Network for Sar Image Classification

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

Transfer Learning with Deep Convolutional Neural Network for SAR Target Classification with Limited Labeled Data

SAR Image Classification Using Greedy Hierarchical Learning With Unsupervised Stacked CAEs

Limited Data-Driven Multi-Task Deep Learning Approach for Target Classification in SAR Imagery

Multiscale CNN with Autoencoder Regularization Joint Contextual Attention Network for SAR Image Classification.

Multi-Scale Convolutional Neural Network for SAR Image Semantic Segmentation

Towards SAR Automatic Target Recognition MultiCategory SAR Image Classification Based on Light Weight Vision Transformer

A Multichannel-Based Deep Learning Framework for Ocean SAR Scene Classification

Synthetic Aperture Radar Target Recognition Of Incomplete Training Datasets Via Siamese Network

High Resolution SAR Image Classification Using Global-Local Network Structure Based on Vision Transformer and CNN

SAR Target Classification Based on Integration of ASC Parts Model and Deep Learning Algorithm

Self-Supervised Classification of SAR Images with Optical Image Assistance