CNN-Enhanced Hypergraph Attention Network for Hyperspectral Image Classification

Ying Cui

doi:10.1109/LGRS.2025.3583697

CNN-Enhanced Hypergraph Attention Network for Hyperspectral Image Classification

2025-01-01

0

PRE

AI

DOI：10.1109/LGRS.2025.3583697

Original

Original request for help

Save

Abstract

En 中文

Recently, graph neural networks (GNNs) have attracted great attention and achieved outstanding success in hyperspectral image (HSI) classification. However, most existing methods rely on pairwise relationship, neglecting more complex higher order interactions, which limits the learning of deeply embedded features. In addition, their dependence on predefined graph structures restricts dynamic node information aggregation. To solve the above problems, this letter proposes a CNN-enhanced hypergraph attention (CEHGAT) network. To reveal the high-order interaction in HSI, a hypergraph attention network branch is developed to learn the dynamic connection of hyperedges through the attention mechanism to reveal more representative node embeddings. Then, the CNN-enhanced branch uses two multiscale convolutional blocks to enhance the spatial-spectral features. Finally, the features captured by two branches are fused to realize the complementary advantages of superpixel-level and pixel-level features. Experiments on three benchmark HSI datasets demonstrate that CEHGAT outperforms other state-of-the-art methods with limited labeled samples.

Keywords:

Convolutional neural network (CNN)

feature fusion

hypergraph attention

hyperspectral image (HSI) classification

CNN-Enhanced Hypergraph Attention Network for Hyperspectral Image Classification

Abstract

Journal

Organization