Abstract: A dynamic graph convolutional network (DGCN) can represent temporal evolutionary features. Its compatibility with the spectral-dimensional characteristics of hyperspectral images (HSIs), ...