Dual Graph Convolutional Network for Hyperspectral Image Classification With Limited Training Samples

Due to powerful feature extraction capability, convolutional neural networks (CNNs) have been widely used for hyperspectral image (HSI) classification. However, because of a large number of parameters that need to be trained, sufficient training samples are usually required for deep CNN-based methods. Unfortunately, limited training samples are a common issue in the remote sensing community. In this study, a dual graph convolutional network (DGCN) is proposed for the supervised classification of HSI with limited training samples. The first GCN fully extracts features existing in and among HSI samples, while the second GCN utilizes label distribution learning, and thus, it potentially reduces the number of required training samples. The two GCNs are integrated through several iterations to decrease interclass distances, which leads to a more accurate classification step. Moreover, a new idea entitled multiscale feature cutout is proposed as a regularization technique for HSI classification (DGCN-M). Different from the regularization methods (e.g., dropout and DropBlock), the proposed multiscale feature cutout could randomly mask out multiscale region sizes in a feature map, which further reduces the overfitting problem and yields consistent improvement. Experimental results on the four popular hyperspectral data sets (i.e., Salinas, Indian Pines, Pavia, and Houston) indicate that the proposed method obtains good classification performance compared to state-of-the-art methods, which shows the potential of GCN for HSI classification.