Summary: Boston University Computer Science Tech. Report No. 2005­009, April 06, 2005.
To appear in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2005.
Efficient Nearest Neighbor Classification Using a Cascade of Approximate
Similarity Measures
Vassilis Athitsos, Jonathan Alon, and Stan Sclaroff
Computer Science Department
Boston University
111 Cummington Street
Boston, MA 02215
email: {athitsos, jalon, sclaroff}@cs.bu.edu
Abstract
This paper proposes a method for efficient nearest neigh­
bor classification in non­Euclidean spaces with computa­
tionally expensive similarity/distance measures. Efficient
approximations of such measures are obtained using the
BoostMap algorithm, which produces embeddings into a
real vector space. A modification to the BoostMap algo­
rithm is proposed, which uses an optimization cost that is
more appropriate when our goal is classification accuracy
as opposed to nearest neighbor retrieval accuracy. Us­

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

Collections: Computer Technologies and Information Sciences