Multiscale Principle of Relevant Information for Hyperspectral Image Classification

TL;DR

This paper introduces MPRI, a multiscale architecture leveraging relevant information principles to improve hyperspectral image classification, especially with limited training data, outperforming existing methods.

Contribution

The paper presents a novel multiscale framework combining relevant information principles with spectral-spatial feature learning for hyperspectral images.

Findings

MPRI outperforms state-of-the-art methods on benchmark datasets.

MPRI is particularly effective with limited training samples.

The approach combines spectral-spatial characterization with dimensionality reduction.

Abstract

This paper proposes a novel architecture, termed multiscale principle of relevant information (MPRI), to learn discriminative spectral-spatial features for hyperspectral image (HSI) classification. MPRI inherits the merits of the principle of relevant information (PRI) to effectively extract multiscale information embedded in the given data, and also takes advantage of the multilayer structure to learn representations in a coarse-to-fine manner. Specifically, MPRI performs spectral-spatial pixel characterization (using PRI) and feature dimensionality reduction (using regularized linear discriminant analysis) iteratively and successively. Extensive experiments on three benchmark data sets demonstrate that MPRI outperforms existing state-of-the-art methods (including deep learning based ones) qualitatively and quantitatively, especially in the scenario of limited training samples. Code of MPRI is available at \url{http://bit.ly/MPRI_HSI}.