Fast Reconstruction of Hyperspectral Image from its RGB Counterpart Using ADMM-Adam Theory

Chia Hsiang Lin, Tzu Hsuan Lin, Ting Hsuan Lin, Tang Huang Lin

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

This paper aims at recovering the hyperspectral image from its RGB counterpart. This highly challenging inverse problem has profoundly impactful applications, including hyperspectral imaging for metamaterial-driven miniaturized satellite. Popular inverse imaging theories include convex optimization (CO, wherein ADMM is a key optimizer) and deep learning (DL, wherein Adam plays a fundamental role); the former usually involves math-heavy optimization procedure, while the latter often requires time-consuming big data collection. We adopt the ADMM-Adam theory, recently investigated in the remote sensing literature for blending the advantages of CO and DL, in order to achieve outstanding hyperspectral signature reconstruction (HSR) without support from heavy math or big data. Simply speaking, a deep regularizer is devised to extract useful information embedded in the rough solution learned from small data. Then, such information is used to design a simple convex regularizer via Q-quadratic function for designing an effective HSR algorithm, whose effectiveness is experimentally illustrated.

Original languageEnglish
Title of host publication2022 12th Workshop on Hyperspectral Imaging and Signal Processing
Subtitle of host publicationEvolution in Remote Sensing, WHISPERS 2022
PublisherIEEE Computer Society
ISBN (Electronic)9781665470698
DOIs
StatePublished - 2022
Event12th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing, WHISPERS 2022 - Rome, Italy
Duration: 13 Sep 202216 Sep 2022

Publication series

NameWorkshop on Hyperspectral Image and Signal Processing, Evolution in Remote Sensing
Volume2022-September
ISSN (Print)2158-6276

Conference

Conference12th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing, WHISPERS 2022
Country/TerritoryItaly
CityRome
Period13/09/2216/09/22

Keywords

  • adaptive moment estimation (Adam)
  • Alternating direction method of multipliers (ADMM)
  • hyperspectral signature reconstruction
  • small data learning

Fingerprint

Dive into the research topics of 'Fast Reconstruction of Hyperspectral Image from its RGB Counterpart Using ADMM-Adam Theory'. Together they form a unique fingerprint.

Cite this