Supervised Spatially Spectrally Coherent Local Linear Embedding—Wasserstein Graph Convolutional Network for Hyperspectral Image Classification

Hyperspectral image (HSI) classification faces challenges due to the high-dimensional and complex nature of the data. Graph convolutional network (GCN) is a widely used method for HSI classification. Conventional GCN analyzes fixed adjacency relations and ignores spatial spectral relationships. To overcome these issues, we propose a supervised spatially spectrally coherent local linear embedding integrated with Wasserstein distance-based GCN ( $\text {S}^{3}$ CLLE-WGCN) for HSI classification. This method combines the strengths of local linear embedding (LLE) for dimensionality reduction (DR) and WD for graph adjacency matrix construction, named Wasserstein GCN (WGCN) for effective feature extraction in non-Euclidean spaces. The supervised spatially spectrally coherent LLE ( $\text {S}^{3}$ CLLE) branch enhances the spectral–spatial manifold representation preserving local geometric variations and class boundaries while effectively reducing the dimensionality of the dataset, while the Earth mover’s or WD better discriminates hyperspectral classes. The $\text {S}^{3}$ CLLE-WGCN achieves state-of-the-art average accuracies: 98.44% (Indian Pines), 99.35% (Houston), 97.92% (Pavia), and 99.74% (Salinas Scene). These results demonstrate that $\text {S}^{3}$ CLLE-WGCN outperforms the baselines in terms of various evaluation criteria.