Deep Amended Gradient Descent for Efficient Spectral Reconstruction from Single RGB Images

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journalpeer-review

15 Scopus Citations
View graph of relations

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)1176-1188
Journal / PublicationIEEE Transactions on Computational Imaging
Volume7
Online published2 Nov 2021
Publication statusPublished - 2021

Abstract

This paper investigates the problem of recovering hyperspectral (HS) images from single RGB images. To tackle such a severely ill-posed problem, we propose a physically-interpretable, compact, efficient and end-to-end learning-based framework, namely AGD-Net. Precisely, by taking advantage of the imaging process, we first formulate the problem explicitly based on the classic gradient descent algorithm. Then, we design a lightweight neural network with a multi-stage architecture to mimic the formed amended gradient descent process, in which efficient convolution and novel spectral zero-mean normalization are proposed to effectively extract spatial-spectral features for regressing an initialization, a basic gradient, and an incremental gradient. Besides, based on the approximate low-rank property of HS images, we propose a novel rank loss to promote the similarity between the global structures of reconstructed and ground-truth HS images, which is optimized with our singular value weighting strategy during training. Moreover, AGD-Net, a single network after one-time training, is flexible to handle the reconstruction with various spectral response functions. Extensive experiments over three commonly-used benchmark datasets demonstrate that AGD-Net can improve the reconstruction quality more than 1.0 dB on average while saving 67x parameters and 32x FLOPs, compared with state-of-the-art methods. The code will be publicly available at https://github.com/zbzhzhy/GD-Net.

Research Area(s)

  • Hyperspectral imagery, spectral reconstruction, deep learning, gradient descent, rank loss