Multilinear sparse decomposition for best spectral bands selection

Hamdi Jamel Bouchech, Sebti Foufou, Mongi Abidi

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Optimal spectral bands selection is a primordial step in multispectral images based systems for face recognition. In this context, we select the best spectral bands using a multilinear sparse decomposition based approach. Multispectral images of 35 subjects presenting 25 different lengths from 480nm to 720nm and three lighting conditions: fluorescent, Halogen and Sun light are groupped in a 3-mode face tensor T of size 35x25x2 . T is then decomposed using 3-mode SVD where three mode matrices for subjects, spectral bands and illuminations are sparsely determined. The 25x25 spectral bands mode matrix defines a sparse vector for each spectral band. Spectral bands having the sparse vectors with the lowest variation with illumination are selected as the best spectral bands. Experiments on two state-of-the-art algorithms, MBLBP and HGPP, showed the effectiveness of our approach for best spectral bands selection.

Original languageEnglish (US)
Title of host publicationImage and Signal Processing - 6th International Conference, ICISP 2014, Proceedings
PublisherSpringer Verlag
Number of pages8
ISBN (Print)9783319079974
StatePublished - 2014
Event6th International Conference on Image and Signal Processing, ICISP 2014 - Cherbourg, France
Duration: Jun 30 2014Jul 2 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8509 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other6th International Conference on Image and Signal Processing, ICISP 2014


  • HGPP
  • Multilinear
  • Spectral bands
  • Tensor
  • sparse

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Multilinear sparse decomposition for best spectral bands selection'. Together they form a unique fingerprint.

Cite this