Self Supervised Learning for Few Shot Hyperspectral Image Classification

TL;DR

This paper introduces a self-supervised learning approach using Barlow-Twins to improve hyperspectral image classification with limited labeled data, significantly outperforming traditional supervised methods.

Contribution

It applies self-supervised learning to hyperspectral image classification, enabling effective model training with minimal labeled data, which is a novel approach in this domain.

Findings

Self-supervised pre-training improves classification accuracy.

The method outperforms traditional supervised learning with fewer labels.

Effective for scenarios with limited labeled hyperspectral data.

Abstract

Deep learning has proven to be a very effective approach for Hyperspectral Image (HSI) classification. However, deep neural networks require large annotated datasets to generalize well. This limits the applicability of deep learning for HSI classification, where manually labelling thousands of pixels for every scene is impractical. In this paper, we propose to leverage Self Supervised Learning (SSL) for HSI classification. We show that by pre-training an encoder on unlabeled pixels using Barlow-Twins, a state-of-the-art SSL algorithm, we can obtain accurate models with a handful of labels. Experimental results demonstrate that this approach significantly outperforms vanilla supervised learning.