Learning Spatial-Spectral Prior for Super-Resolution of Hyperspectral Imagery

TL;DR

This paper introduces a novel spatial-spectral prior network (SSPN) for hyperspectral image super-resolution, effectively leveraging spectral and spatial correlations to improve detail recovery, while addressing training stability with group convolution and progressive upsampling.

Contribution

It adapts residual learning-based super-resolution techniques for hyperspectral images using a specialized network that exploits spatial-spectral priors and stabilizes training with group convolution.

Findings

Outperforms existing methods in hyperspectral super-resolution

Enhances details in high-resolution hyperspectral images

Demonstrates stable training with high-dimensional data

Abstract

Recently, single gray/RGB image super-resolution reconstruction task has been extensively studied and made significant progress by leveraging the advanced machine learning techniques based on deep convolutional neural networks (DCNNs). However, there has been limited technical development focusing on single hyperspectral image super-resolution due to the high-dimensional and complex spectral patterns in hyperspectral image. In this paper, we make a step forward by investigating how to adapt state-of-the-art residual learning based single gray/RGB image super-resolution approaches for computationally efficient single hyperspectral image super-resolution, referred as SSPSR. Specifically, we introduce a spatial-spectral prior network (SSPN) to fully exploit the spatial information and the correlation between the spectra of the hyperspectral data. Considering that the hyperspectral training samples are scarce and the spectral dimension of hyperspectral image data is very high, it is nontrivial to train a stable and effective deep network. Therefore, a group convolution (with shared network parameters) and progressive upsampling framework is proposed. This will not only alleviate the difficulty in feature extraction due to high-dimension of the hyperspectral data, but also make the training process more stable. To exploit the spatial and spectral prior, we design a spatial-spectral block (SSB), which consists of a spatial residual module and a spectral attention residual module. Experimental results on some hyperspectral images demonstrate that the proposed SSPSR method enhances the details of the recovered high-resolution hyperspectral images, and outperforms state-of-the-arts. The source code is available at \url{https://github.com/junjun-jiang/SSPSR