A Semi-supervised SAR Image Classification with Data Augmentation and Pseudo Labeling

We present a semi-supervised Synthetic Aperture Radar (SAR) image classification method to improve the SAR classification performance with a small number of datasets. To overcome the disadvantage of a small number of SAR data, we apply FixMatch method to SAR image dataset. We further augment data by adding speckle noise to handle the speckle noise in SAR image. By leveraging the augmented data as strongly augmented data in FixMatch method, we effectively learn the network with small amount of labelled data. Experimental results demonstrated that the proposed method outperforms baseline.