Interest Point Stability Prediction

Authors:
H. Thomson Comer;Bruce A. Draper
Affiliations:
Computer Science Department, Colorado State University, USA 80523;Computer Science Department, Colorado State University, USA 80523
Venue:
ICVS '09 Proceedings of the 7th International Conference on Computer Vision Systems: Computer Vision Systems
Year:
2009

Citing 8
Cited 0

Feature Detection with Automatic Scale Selection

International Journal of Computer Vision
Saliency, Scale and Image Description

International Journal of Computer Vision
An Affine Invariant Interest Point Detector

ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part I
Scale & Affine Invariant Interest Point Detectors

International Journal of Computer Vision
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Local Features for Object Class Recognition

ICCV '05 Proceedings of the Tenth IEEE International Conference on Computer Vision - Volume 2
A Comparison of Affine Region Detectors

International Journal of Computer Vision
Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories

Computer Vision and Image Understanding

Quantified Score

Hi-index	0.00

Visualization

Abstract

Selective attention algorithms produce more interest points than are usable by many computer vision applications. This work suggests a method for ordering and selecting a subset of those interest points, simultaneously increasing the repeatability of that subset. The individual repeatability of a combination of 106 SIFT, Harris-Laplace, and Hessian-Laplace interest points are predicted using generalized linear models (GLMs). The models are produced by studying the 17 attributes of each interest point. Our goal is not to improve any particular algorithm, but to find attributes that affect affine and similarity invariance regardless of algorithm. The techniques explored in this research enable interest point detectors to improve mean repeatability of their algorithm by 4% using a rank-ordering produced by a GLM or by thresholding interest points using a set of five new thresholds. Selecting the top 1% of GLM-ranked Harris-Laplace interest points results in a repeatability improvement of 6%, to 92.4%.