Weighted residual network for SAR automatic target recognition with data augmentation.
Journal:
Frontiers in neurorobotics
Published Date:
Dec 19, 2023
Abstract
INTRODUCTION: Decades of research have been dedicated to overcoming the obstacles inherent in synthetic aperture radar (SAR) automatic target recognition (ATR). The rise of deep learning technologies has brought a wave of new possibilities, demonstrating significant progress in the field. However, challenges like the susceptibility of SAR images to noise, the requirement for large-scale training datasets, and the often protracted duration of model training still persist.
Authors
Keywords
No keywords available for this article.