Match-time covariance for descriptors
Research output: Contribution to conference › Paper › Research › peer-review
Standard
Match-time covariance for descriptors. / Christiansen, Eric; Rabaud, Vincent; Ziegler, Andrew; Kriegman, David; Belongie, Serge.
2013. Paper presented at 2013 24th British Machine Vision Conference, BMVC 2013, Bristol, United Kingdom.Research output: Contribution to conference › Paper › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - CONF
T1 - Match-time covariance for descriptors
AU - Christiansen, Eric
AU - Rabaud, Vincent
AU - Ziegler, Andrew
AU - Kriegman, David
AU - Belongie, Serge
PY - 2013
Y1 - 2013
N2 - Local descriptor methods are widely used in computer vision to compare local regions of images. These descriptors are often extracted relative to an estimated scale and rotation to provide invariance up to similarity transformations. The estimation of rotation and scale in local neighborhoods (also known as steering) is an imperfect process, however, and can produce errors downstream. In this paper, we propose an alternative to steering that we refer to as match-time covariance (MTC). MTC is a general strategy for descriptor design that simultaneously provides invariance in local neighborhood matches together with the associated aligning transformations. We also provide a general framework for endowing existing descriptors with similarity invariance through MTC. The framework, Similarity-MTC, is simple and dramatically improves accuracy. Finally, we propose NCC-S, a highly effective descriptor based on classic normalized cross-correlation, designed for fast execution in the Similarity-MTC framework. The surprising effectiveness of this very simple descriptor suggests that MTC offers fruitful research directions for image matching previously not accessible in the steering based paradigm.
AB - Local descriptor methods are widely used in computer vision to compare local regions of images. These descriptors are often extracted relative to an estimated scale and rotation to provide invariance up to similarity transformations. The estimation of rotation and scale in local neighborhoods (also known as steering) is an imperfect process, however, and can produce errors downstream. In this paper, we propose an alternative to steering that we refer to as match-time covariance (MTC). MTC is a general strategy for descriptor design that simultaneously provides invariance in local neighborhood matches together with the associated aligning transformations. We also provide a general framework for endowing existing descriptors with similarity invariance through MTC. The framework, Similarity-MTC, is simple and dramatically improves accuracy. Finally, we propose NCC-S, a highly effective descriptor based on classic normalized cross-correlation, designed for fast execution in the Similarity-MTC framework. The surprising effectiveness of this very simple descriptor suggests that MTC offers fruitful research directions for image matching previously not accessible in the steering based paradigm.
UR - http://www.scopus.com/inward/record.url?scp=84898491145&partnerID=8YFLogxK
U2 - 10.5244/C.27.12
DO - 10.5244/C.27.12
M3 - Paper
AN - SCOPUS:84898491145
T2 - 2013 24th British Machine Vision Conference, BMVC 2013
Y2 - 9 September 2013 through 13 September 2013
ER -
ID: 302046402