Fusing Spatial Attention with Spectral-Channel Attention Mechanism for Hyperspectral Image Classification via Encoder–Decoder Networks

In recent years, convolutional neural networks (CNNs) have been widely used in hyperspectral image (HSI) classification. However, feature extraction on data still faces numerous challenges. Existing methods cannot extract spatial and spectral-channel contextual information a targeted manner. this paper, we propose an encoder–decoder network that fuses attention for HSI classification from three public datasets to tackle these issues. terms of fusion, multi-source mechanism including is proposed encode the spectral multi-channels information. Moreover, fusion strategies are effectively utilize attention. They direct aggregation, aggregation space, Hadamard product. development, framework employed The encoder hierarchical transformer pipeline can long-range context Both shallow local features rich global semantic encoded through expressions. decoder consists suitable upsampling, skip connection, convolution blocks, which fuse multi-scale efficiently. Compared with other state-of-the-art methods, our approach has greater performance