Closest Point Search in High Dimensions

Closest Point Search in High Dimensions (1996)

Cached

Download Links

[ftp.cs.columbia.edu]

Other Repositories/Bibliography

DBLP

by Sameer A. Nene , Shree K. Nayar

Venue:	Proc. of IEEE Conference on Computer Vision and Pattern Recognition, (CVPR 96
Citations:	10 - 0 self

Versions

Version History

Metadata Version 2

Datum	Value	Source
TITLE	Closest Point Search in High Dimensions	SVM HeaderParse 0.1
AUTHOR NAME	Sameer A. Nene	SVM HeaderParse 0.1
AUTHOR AFFIL	Department of Computer Science, Columbia University	SVM HeaderParse 0.2
AUTHOR ADDR	New York, NY 10027	SVM HeaderParse 0.1
AUTHOR NAME	Shree K. Nayar	SVM HeaderParse 0.1
AUTHOR AFFIL	Department of Computer Science, Columbia University	SVM HeaderParse 0.2
AUTHOR ADDR	New York, NY 10027	SVM HeaderParse 0.1
ABSTRACT	The problem of finding the closest point in highdimensional spaces is common in computational vision. Unfortunately, the complexity of most existing search algorithms, such as k-d tree and R-tree, grows exponentially with dimension, making them impractical for dimensionality above 15. In nearly all applications, the closest point is of interest only if it lies within a user specified distance ffl. We present a simple and practical algorithm to efficiently search for the nearest neighbor within Euclidean distance ffl. Our algorithm uses a projection search technique along with a novel data structure to dramatically improve performance in high dimensions. A complexity analysis is presented which can help determine ffl in structured problems. Benchmarks clearly show the superiority of the proposed algorithm for high dimensional search problems frequently encountered in machine vision, such as real-time object recognition.	SVM HeaderParse 0.1
YEAR	1996	INFERENCE
VENUE	Proc. of IEEE Conference on Computer Vision and Pattern Recognition, (CVPR 96	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	859--865	INFERENCE
VOLUME	96	INFERENCE
CITATIONS	13 found	ParsCit 1.0