Multiclass feature learning for hyperspectral image classification: sparse and hierarchical solutions

TL;DR

This paper introduces an active set feature learning approach for hyperspectral image classification that automatically discovers effective spatial filters, including hierarchical features, enabling simple classifiers to achieve state-of-the-art performance.

Contribution

The paper proposes a novel active set feature learning method that automatically identifies spatial filters for hyperspectral classification, incorporating hierarchical features for improved modeling.

Findings

Achieved state-of-the-art classification performance on multiple hyperspectral datasets.

Demonstrated the effectiveness of hierarchical feature banks in capturing data nonlinearities.

Validated the approach across agricultural, urban, and multimodal hyperspectral data.

Abstract

In this paper, we tackle the question of discovering an effective set of spatial filters to solve hyperspectral classification problems. Instead of fixing a priori the filters and their parameters using expert knowledge, we let the model find them within random draws in the (possibly infinite) space of possible filters. We define an active set feature learner that includes in the model only features that improve the classifier. To this end, we consider a fast and linear classifier, multiclass logistic classification, and show that with a good representation (the filters discovered), such a simple classifier can reach at least state of the art performances. We apply the proposed active set learner in four hyperspectral image classification problems, including agricultural and urban classification at different resolutions, as well as multimodal data. We also propose a hierarchical setting, which allows to generate more complex banks of features that can better describe the nonlinearities present in the data.