Repository logo

Infoscience

  • English
  • French
Log In
Logo EPFL, École polytechnique fédérale de Lausanne

Infoscience

  • English
  • French
Log In
  1. Home
  2. Academic and Research Output
  3. Reports, Documentation, and Standards
  4. Decision fusion in a multi-modal identity verification system using a multi-linear classifier
 
report

Decision fusion in a multi-modal identity verification system using a multi-linear classifier

Verlinde, Patrick
•
Maître, Gilbert
•
Mayoraz, Eddy
1997

This paper presents the use of a multi-linear classifier allowing to fuse the results of several modalities in a multi-modal person identity verification context. In the considered verification system, each of the d modalities forms an autonomous bloc that produces a score, which is not only supposed to be monotone but also to have a value between zero and one. The fusion module that we are discussing here takes a binary decision: accept or reject the identity claimed by the person, based on the whole of the scores given in parallel by all d modalities. To realize this fusion module we have developed a classifier that, on the one hand, accepts the monotonicity hypothesis and, on the other hand, is based on separating the classes (accept-reject) by a combination of half-spaces, a technique from which it derived its name. The classifier is trained using couples formed by extracting an example from each class and the half-spaces are determined by maximizing a global separability measure of the thus formed couples. Afterwards, each region of the partition of the d dimensional space, generated by the half-spaces, is labeled with the corresponding class, using the Logical Analysis of Data (LAD) method. The performance of the developed multi-linear classifier has been evaluated on multi-modal experimental data and the obtained results are presented.

  • Details
  • Metrics
Type
report
Author(s)
Verlinde, Patrick
Maître, Gilbert
Mayoraz, Eddy
Date Issued

1997

Publisher

IDIAP

Subjects

vision

•

learning

•

eddy

Written at

EPFL

EPFL units
LIDIAP  
Available on Infoscience
March 10, 2006
Use this identifier to reference this record
https://infoscience.epfl.ch/handle/20.500.14299/227750
Logo EPFL, École polytechnique fédérale de Lausanne
  • Contact
  • infoscience@epfl.ch

  • Follow us on Facebook
  • Follow us on Instagram
  • Follow us on LinkedIn
  • Follow us on X
  • Follow us on Youtube
AccessibilityLegal noticePrivacy policyCookie settingsEnd User AgreementGet helpFeedback

Infoscience is a service managed and provided by the Library and IT Services of EPFL. © EPFL, tous droits réservés