In this paper, we present Concurrent Self-Organizing Maps (CSOM), a new artificial neural classification model representing a winner-takes-all collection of small SOM units. We consider two significant areas of CSOM applications in Biometric Technology: face recognition and speaker recognition. For the ORL face database of 40 subjects, CSOM yields a recognition score of 91%, while a single, large SOM yields a score of only 71%! For a speaker database provided by 25 talkers, a recognition score of 92.17% was obtained using CSOM, compared to the recognition rate of 79.63% yielded by the SOM. This model may be applied in access control applications for harbour protection.
Concurrent Self-Organizing Maps — A Powerful Artificial Neural Tool for Biometric Technology
01.01.2009
8 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Concurrent Self-Organizing Maps , biometric technology , face recognition , speaker recognition Computer Science , Computer Systems Organization and Communication Networks , Applications of Mathematics , Information Systems and Communication Service , Computer Imaging, Vision, Pattern Recognition and Graphics , Computer Appl. in Social and Behavioral Sciences
Concurrent Self-Organizing Maps - A Powerful Artificial Neural Tool for Biometric Technology
British Library Conference Proceedings | 2009
|Using Artificial Neural Networks and Self-Organizing Maps for Detection of Airframe Icing
British Library Conference Proceedings | 2000
|