DocumentCode
3525312
Title
Boosted ECOC ensembles for face recognition
Author
Windeatt, T. ; Ardeshir, G.
Author_Institution
Surrey Univ., Guildford, UK
fYear
2003
fDate
7-9 July 2003
Firstpage
165
Lastpage
168
Abstract
The error correcting output coding (ECOC) approach to classifier design decomposes a multi-class problem into a set of complementary two-class problems. In this paper ECOC is combined with boosting using decision tree base classifiers to solve two problems in face recognition. Boosted ECOC ensembles using decision stumps and C4.5 classifiers are compared for face identification and verification.
Keywords
computer vision; decision trees; error correction codes; face recognition; visual databases; Adaboost; boosting; decision stumps; decision tree; error correcting output coding; face database; face recognition; face representation;
fLanguage
English
Publisher
iet
Conference_Titel
Visual Information Engineering, 2003. VIE 2003. International Conference on
ISSN
0537-9989
Print_ISBN
0-85296-757-8
Type
conf
DOI
10.1049/cp:20030513
Filename
1341318
Link To Document