Our experiments demonstrate that this model outperforms both pre-trained and scratch DCNN models in terms of classification accuracy for augmented data. Specifically, our approach shows superior ...

This article introduces a two-phase learning approach for hyperspectral image (HSI) classification using few-shot learning (FSL). For the first phase, we present a novel spatiospectral masked ...