Pseudo Complex-Valued Deformable ConvLSTM Neural Network With Mutual Attention Learning for Hyperspectral Image Classification

Convolutional long short-term memory (ConvLSTM) has received much attention for hyperspectral image (HSI) classification due to its ability of modeling long-range correlations, which, however, is vulnerable to too many parameters and insufficient training, limiting its classification accuracy, especially for small samples. Different from it, traditional hand-crafted methods extract the features with basic attributes of HSIs, which can provide the lack of details and interpretability of deep semantic features. However, existing methods fail to incorporate their complementarity for HSI classification. As such, a Pseudo complex-valued (CV) Deformable ConvLSTM Neural Network with mutual Attention learning (APDCLNN) is proposed, providing a new way to realize the collaborative learning of hand-crafted and deep features for HSI classification. First, a 2-D pseudo CV deformable ConvLSTM (PDConvLSTM2D) cell is designed using deformable convolution and complex operations, with which a spatial–spectral PDConvLSTM2D neural network (SSPDCL2DNN) is built to extract scale- and spectral-enhanced deep spatial–spectral features. Then, 3-D Gabor filter is used to extract hand-crafted features, and a mutual attention-based multimodality feature learning and fusion (MAMLF) module is designed to integrate them into deep features for training and optimization of SSPDCL2DNN. Finally, an attention loss subnetwork is designed to refine the classification results. As we know, this is the first attempt to apply the idea of mutual attention learning to fuse hand-crafted and deep features for HSI classification. Extensive experiments on three widely used HSI datasets show the advantages of our model over other deep methods in terms of both quantitative and visual quality.