Geunhyuk Youk' Paper Accepted for IEEE TGRS Journal

Congratulations!

Geunhyuk Youk' Paper has been accepted for IEEE Transactions on Geoscience and Remote Sensing (IF 8.125).

Title: Transformer-based Synthetic-to-Measured SAR Image Translation via Learning of Representational Features

Authos: Geunhyuk Youk and Munchurl Kim

Abstract:

Deep learning-based target recognition in synthetic aperture radar (SAR) images has been actively studied in recent years. However, it is very costly to collect large amounts of labeled SAR images, especially measured SAR target images of various classes, to train high-performance classification networks. To solve the problem of insufficient SAR data, electromagnetic computational tools have often been developed and used to synthesize the measure SAR target images from data modeling. However, despite of sophisticated SAR image modeling, there is a large domain gap between synthetic SAR images and measured ones, so that the networks trained with synthetic SAR images tend to show poor classification performance when tested for measured SAR target images. In this paper, we propose a novel transformer-based synthetic-to-measured SAR target image translation network, called SAR-SMT Net, to bridge the gap between synthetic and measured SAR target images. The SAR-SMT Net takes synthetic SAR target images as input, and estimates the latent representational features of their corresponding measured SAR ones to faithfully adjust the global context and scattering characteristics of the input synthetic SAR target images to their corresponding measured SAR ones. In addition, we propose five challenging experiment scenarios that can validate the performance of SAR image translation. Experimentally, our SAR-SMT Net outperforms previous state-of-the-art methods in the experiment scenarios, and shows a well generalization ability in translating the synthetic SAR target images to their corresponding measured SAR target ones with a high level of fidelity even for unseen target classes at unseen azimuth angles.