A Comparison of Deep Learning Algorithms Dealing With Limited Samples in Hyperspectral Image Classification

Abstract

Hyperspectral Imaging, also known as Spectroscopy, is used in various areas such as medicine, defense, submarine, remote sensing, and environmental monitoring. Several supervised or unsupervised deep learning algorithms have been developed to classify such hyperspectral images. A significant problem in HSI is insufficient data availability, as annotating the samples is time-consuming and labor-intensive. This study provides a comparison of deep learning algorithms that have been developed to deal with the limited data problem in the HSI domain. It compares the performance, classification accuracy and other relevant parameters that exist during the development of such algorithms.