Blind Hyperspectral Unmixing using Autoencoders: A Critical Comparison

Research output: Contribution to journalArticlepeer-review


Deep learning has shown to be a powerful tool and has heavily impacted the data-intensive field of remote sensing. As a result, the number of published deep learning-based spectral unmixing techniques is proliferating. Blind hyperspectral unmixing (HU) is the process of resolving the measured spectrum of a pixel into a combination of a set of spectral signatures called endmembers and simultaneously determining their fractional abundances in the pixel. This paper details the various autoencoder architectures used in HU and provides a critical comparison of some of the existing published blind unmixing methods based on autoencoders. Eleven different autoencoder methods and one traditional method will be compared in blind unmixing experiments using four real datasets and four synthetic datasets with different spectral variability. Additionally, extensive ablation experiments with a simple spectral unmixing autoencoder will be performed. The results are interpreted in terms of the various implementation details, and the question of why autoencoder methods are so powerful compared to traditional methods is unraveled. The source codes for all methods implemented in this paper can be found at the following location:

Original languageEnglish
Pages (from-to)1340-1372
Number of pages33
JournalIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Publication statusPublished - 1 Jan 2022

Bibliographical note

Funding Information:
This work was supported in part by the Icelandic Research Fund under Grant 174075-05 and 207233-052.

Publisher Copyright:
© 2008-2012 IEEE.

Other keywords

  • Atmospheric modeling
  • autoencoder
  • deep learning
  • Feature extraction
  • Hyperspectral data unmixing
  • Hyperspectral imaging
  • image processing
  • multitask learning
  • neural network
  • Object detection
  • Sparse matrices
  • Spatial resolution
  • spectral-spatial model
  • Spectroscopy


Dive into the research topics of 'Blind Hyperspectral Unmixing using Autoencoders: A Critical Comparison'. Together they form a unique fingerprint.

Cite this