Efficient Unsupervised Deep Band Selection for Hyperspectral Imagery: A Comparative Study With Mamba-Based Classification

Abstract

Band selection is a critical step in processing hyperspectral imagery (HSI); reducing input dimensionality allows models to mitigate redundancy, enhance computational efficiency, and improve learning accuracy. Efficient unsupervised deep-learning-based band selection methods have recently garnered immense attention due to their advanced feature representation capabilities. The existing literature shows there is a broader and more general line of research regarding feature selection, from which some recent deep learning-based HSI band selection methods have drawn inspiration. This work is a comparative study focusing on efficient unsupervised deep-learning-based HSI band selection methods, most of which are adapted from the general feature selection. A benchmarking analysis was conducted in terms of downstream classification performance and computation cost, on six state-of-the-art efficient unsupervised HSI band selection methods. Classification experiments were carried out using three publicly available remote sensing datasets and three classifiers—MambaHSI, HyPyraMamba, and a Support Vector Machine. This work provides a unified comparison of these efficient unsupervised methods for HSI band selection and investigates the use of Mamba-based classification in this context.