A novel discriminant non-negative matrix factorization algorithm with applications to facial image characterization problems

Article


Kotsia, I., Zafeiriou, S. and Pitas, I. 2007. A novel discriminant non-negative matrix factorization algorithm with applications to facial image characterization problems. IEEE Transactions on Information Forensics and Security. 2 (3), pp. 588-595.
TypeArticle
TitleA novel discriminant non-negative matrix factorization algorithm with applications to facial image characterization problems
AuthorsKotsia, I., Zafeiriou, S. and Pitas, I.
Abstract

The methods introduced so far regarding discriminant non-negative matrix factorization (DNMF) do not guarantee convergence to a stationary limit point. In order to remedy this limitation, a novel DNMF method is presented that uses projected gradients. The proposed algorithm employs some extra modifications that make the method more suitable for classification tasks. The usefulness of the proposed technique to frontal face verification and facial expression recognition problems is demonstrated.

Research GroupResearch Group on Development of Intelligent Environments
PublisherInstitute of Electrical and Electronics Engineers
JournalIEEE Transactions on Information Forensics and Security
ISSN1556-6013
Publication process dates
Deposited26 Nov 2012
Output statusPublished
Web address (URL)http://dx.doi.org/10.1109/TIFS.2007.902017
LanguageEnglish
Permalink -

https://repository.mdx.ac.uk/item/83wv7

  • 22
    total views
  • 0
    total downloads
  • 0
    views this month
  • 0
    downloads this month

Export as