Composite kernel learning network for hyperspectral image classification

The small sample problem has always been a serious challenge in hyperspectral image (HSI) classification. In order to obtain satisfactory results when the training samples are insufficient, the inf...