Summary: To appear in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2009.
Reducing JointBoost­Based Multiclass Classification to Proximity Search
Alexandra Stefan, Vassilis Athitsos
Computer Science and Engineering Department
University of Texas at Arlington
Quan Yuan, Stan Sclaroff
Computer Science Department
Boston University
Abstract
Boosted one­versus­all (OVA) classifiers are commonly
used in multiclass problems, such as generic object recog­
nition, biometrics­based identification, or gesture recogni­
tion. JointBoost is a recently proposed method where OVA
classifiers are trained jointly and are forced to share fea­
tures. JointBoost has been demonstrated to lead both to
higher accuracy and smaller classification time, compared
to using OVA classifiers that were trained independently
and without sharing features. However, even with the im­
proved efficiency of JointBoost, the time complexity of OVA­
based multiclass recognition is still linear to the number of

Source: Athitsos, Vassilis - Department of Computer Science and Engineering, University of Texas at Arlington

Collections: Computer Technologies and Information Sciences