Deep learning-based spectral band selection for hyperspectral imaging tasks

Abstract

An end-to-end framework that jointly learns task-specific hyperspectral band selection and a neural model for classification and segmentation.

Publication
In Applied Optics, 65(13), F75–F85