Abstract: Graph Neural Networks (GNNs) have emerged as a promising solution for few-shot hyperspectral image (HSI) classification. However, existing GNN-based approaches face critical limitations in ...
Abstract: The problem solved in the article is reduction of computational costs for the image classification process when applying structural methods. The main focus is implementing tools for ...