DACN: Dual-Attention Convolutional Network for Hyperspectral Image Super-Resolution

TL;DR

This paper introduces DACN, a dual-attention convolutional network that combines multi-head and channel attention mechanisms to improve hyperspectral image super-resolution by capturing both local and global features.

Contribution

The novel dual-attention architecture effectively integrates multi-head and channel attention for enhanced hyperspectral image super-resolution performance.

Findings

Outperforms existing methods on two hyperspectral datasets.

Multi-head and channel attention combination yields superior results.

Proposed loss function improves spectral fidelity.

Abstract

2D convolutional neural networks (CNNs) have attracted significant attention for hyperspectral image super-resolution tasks. However, a key limitation is their reliance on local neighborhoods, which leads to a lack of global contextual understanding. Moreover, band correlation and data scarcity continue to limit their performance. To mitigate these issues, we introduce DACN, a dual-attention convolutional network for hyperspectral image super-resolution. Specifically, the model first employs augmented convolutions, integrating multi-head attention to effectively capture both local and global feature dependencies. Next, we infer separate attention maps for the channel and spatial dimensions to determine where to focus across different channels and spatial positions. Furthermore, a custom optimized loss function is proposed that combines L2 regularization with spatial-spectral gradient loss to ensure accurate spectral fidelity. Experimental results on two hyperspectral datasets demonstrate that the combination of multi-head attention and channel attention outperforms either attention mechanism used individually.