Hyperspectral Image Few-Shot Classification Network Based on the Earth Mover's Distance

Deep learning has achieved promising performance in hyperspectral image (HSI) classification. Training deep models usually requires labeling massive HSIs, which, however, is prohibitively time-consuming and expensive. To fill in the gap, this article proposes a novel meta-learning method for HSI few-shot classification that conducts HSI classification with a few labeled samples. Specifically, we introduce the Earth mover's distance (EMD) as a metric. The designed EMD metric learning module aims to calculate the similarity of paired embedding features by decomposing embedding features into a set of local representations. The EMD metric aims to find the optimal matching flows between local representations that have the minimum matching cost. Furthermore, we attempt to learn class prototype representation for each hyperspectral class using the EMD metric. The proposed network effectively learns general knowledge from base HSIs and transfers such knowledge to the classification of novel HSIs. We conduct HSI few-shot classification by training on three base HSIs and classification on three novel HSIs. Extensive experimental results on three novel HSI datasets demonstrate that the proposed model outperforms the existing state-of-the-art HSI methods, including two HSI few-shot methods.

Automated summary

This article proposes a novel meta-learning method for hyperspectral image (HSI) few-shot classification. It utilizes a few labeled samples for classification. The Earth mover's distance (EMD) is introduced as a metric, and the EMD metric learning module is designed to calculate the similarity of embedding features. The proposed method outperforms existing HSI methods according to extensive experimental results.