Grouped Multi-Attention Network for Hyperspectral Image Spectral-Spatial Classification

Deep learning (DL) has been a powerful tool for hyperspectral image (HSI) classification. However, it is still an open issue to effectively learn highly discriminative features from HSI, due to the high-dimensionality and complex spectral-spatial characteristics. To settle this issue, we propose a new band-grouping guided multi-attention module for the performance promotion of spectral-spatial feature learning. First, based on the fact of high relevance between adjacent spectral bands and low dependencies across long-range ones, all the spectral bands are adaptively divided into multiple nonoverlapping groups where relevant bands are included. The advantage is to reduce the spectral dimension and data complexity when processing and analyzing each group. Then, a multi-attention mechanism, which not only explores the intragroup salient information but also propagates the intergroup difference information, is embedded into the convolutional neural networks (CNNs) to learn group-specific spectral-spatial features. By emphasizing useful spectral/spatial information and squeezing useless information with attention mechanism, the severability of learned features is enhanced. Based on this module, a spectral-spatial classification network is built, named by grouped multi-attention network (GMA-Net). The GMA-Net contains a two-branch architecture, i.e., pixelwise spectral feature learning and patchwise spectral-spatial feature learning. Via fusing the features from two branches, the complementary and discriminative features provided by pixelwise and patchwise learning manner can be integrated to further boost the classification performance. Experimental results demonstrate that the proposed method is superior than several state-of-the-art approaches. Codes are available at: https://github.com/luting-hnu.

Automated summary

This study proposes a new band-grouping guided multi-attention module to enhance the performance of spectral-spatial feature learning. Spectral bands are adaptively divided into multiple nonoverlapping groups, reducing complexity. A multi-attention mechanism is embedded into CNNs to learn group-specific spectral-spatial features. A spectral-spatial classification network is built, integrating pixelwise and patchwise learning to boost performance.