2 - A brief introduction to deep learning

Abstract

This chapter introduces Deep Learning (DL) as a transformative paradigm in remote sensing, particularly in the context of Synthetic Aperture Radar (SAR) data. While traditional methods — from manual interpretation to physically inspired models — remain important, DL is increasingly integrated into operational workflows. The chapter offers a high-level overview of core DL concepts as a foundation for later chapters focused on SAR applications. Supported by open datasets and tools, DL is rapidly advancing applied SAR domains including disaster response, climate monitoring, and land use analysis.