Disentangling the Benefits of Self-Supervised Learning to Deployment-Driven Downstream Tasks of Satellite Images (Student Abstract)

In this paper, we investigate the benefits of self-supervised learning (SSL) to downstream tasks satellite images. Unlike common student academic projects, work focuses on advantages SSL for deployment-driven which have specific scenarios with low or high-spatial resolution Our preliminary experiments demonstrate robust trained by medium-resolution (10m) images both low-resolution (100m) scene classification case (4.25%↑) and very high-resolution (5cm) aerial image segmentation (1.96%↑), respectively.