In recent years, with the emergence of new technologies for deep learning, deep learning is widely used for hyperspectral image classification. convolutional neural network is one of the most frequently used deep learning based methods for visual data processing. The use of CNN for hyperspectral image classification is also visible in recent works. However, the classification effect is not satisfactory when limited training samples are available. Focused on “small sample” hyperspectral classification, we proposed a novel 3D-2D-convolutional neural network model named SE-HybridSN. In SE-HybridSN model, a dense block was used to reuse shallow features and aimed at better exploiting hierarchical spatial–spectral features. Subsequent depth separable convolutional layers were used to discriminate the spatial information. Further refinement of spatial– spectral features was realized by the channel attention method, which were performed behind every 3D convolutional layer and every 2D convolutional layer. Experiment results indicate that our proposed model can learn more discriminative spatial–spectral features using very few training data. In Indian Pines, Salinas and the University of Pavia, SE-HybridSN using only 5%, 1% and 1%labeled data for training. A very satisfactory performance is obtained using the proposed SE-HybridSN.