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 Using Contrastive Learning and Pseudo-Labels with Limited Data

Synthetic Aperture Radar Image Classification Based on Constrictive Learning with Limited Data

Self-Supervised Classification of SAR Images with Optical Image Assistance

SAR Image Interpretation Using CNN and Contrastive Learning

Self-Supervised Feature Representation for SAR Image Target Classification Using Contrastive Learning

Self-Supervised Contrastive Learning on Cross-Augmented Samples for SAR Target Recognition

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

Remote Sensing Image Classification with Few Labeled Data Using Semisupervised Learning

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

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

Sar target classification with limited data via data driven active learning

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

Locality Preserving Property Constrained Contrastive Learning for Object Classification in SAR Imagery

Multi-scale contrastive learning method for PolSAR image classification

SAR Target Recognition Using Only Simulated Data for Training by Hierarchically Combining CNN and Image Similarity.

Deep Supervised and Contractive Neural Network for Sar Image Classification

Polarimetric SAR Images Classification Based on Convolutional Neural Networks.

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

A Semi-Supervised Synthetic Aperture Radar (SAR) Image Recognition Algorithm Based on an Attention Mechanism and Bias-Variance Decomposition

Contrastive Learning-Based Dual Dynamic GCN for SAR Image Scene Classification.

Semi-Supervised SAR ATR Based on Contrastive Learning and Complementary Label Learning