Robust near-isometric matching via structured learning of graphical models

Julian J. McAuley; Tibério S. Caetano; Alexander J. Smola

Robust near-isometric matching via structured learning of graphical models

Julian J. McAuley^*, Tibério S. Caetano, Alexander J. Smola

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

Models for near-rigid shape matching are typically based on distance-related features, in order to infer matches that are consistent with the isometric assumption. However, real shapes from image datasets, even when expected to be related by "almost isometric" transformations, are actually subject not only to noise but also, to some limited degree, to variations in appearance and scale. In this paper, we introduce a graphical model that parameterises appearance, distance, and angle features and we learn all of the involved parameters via structured prediction. The outcome is a model for near-rigid shape matching which is robust in the sense that it is able to capture the possibly limited but still important scale and appearance variations. Our experimental results reveal substantial improvements upon recent successful models, while maintaining similar running times.

Original language	English
Title of host publication	Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference
Publisher	Neural Information Processing Systems
Pages	1057-1064
Number of pages	8
ISBN (Print)	9781605609492
Publication status	Published - 2009
Event	22nd Annual Conference on Neural Information Processing Systems, NIPS 2008 - Vancouver, BC, Canada Duration: 8 Dec 2008 → 11 Dec 2008

Publication series

Name	Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference

Conference

Conference	22nd Annual Conference on Neural Information Processing Systems, NIPS 2008
Country/Territory	Canada
City	Vancouver, BC
Period	8/12/08 → 11/12/08

Cite this

McAuley, J. J., Caetano, T. S., & Smola, A. J. (2009). Robust near-isometric matching via structured learning of graphical models. In Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference (pp. 1057-1064). (Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference). Neural Information Processing Systems.

McAuley, Julian J. ; Caetano, Tibério S. ; Smola, Alexander J. / Robust near-isometric matching via structured learning of graphical models. Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference. Neural Information Processing Systems, 2009. pp. 1057-1064 (Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference).

@inproceedings{0f253755c81148598297ace13db67879,

title = "Robust near-isometric matching via structured learning of graphical models",

abstract = "Models for near-rigid shape matching are typically based on distance-related features, in order to infer matches that are consistent with the isometric assumption. However, real shapes from image datasets, even when expected to be related by {"}almost isometric{"} transformations, are actually subject not only to noise but also, to some limited degree, to variations in appearance and scale. In this paper, we introduce a graphical model that parameterises appearance, distance, and angle features and we learn all of the involved parameters via structured prediction. The outcome is a model for near-rigid shape matching which is robust in the sense that it is able to capture the possibly limited but still important scale and appearance variations. Our experimental results reveal substantial improvements upon recent successful models, while maintaining similar running times.",

author = "McAuley, {Julian J.} and Caetano, {Tib{\'e}rio S.} and Smola, {Alexander J.}",

year = "2009",

language = "English",

isbn = "9781605609492",

series = "Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference",

publisher = "Neural Information Processing Systems",

pages = "1057--1064",

booktitle = "Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference",

note = "22nd Annual Conference on Neural Information Processing Systems, NIPS 2008 ; Conference date: 08-12-2008 Through 11-12-2008",

}

McAuley, JJ, Caetano, TS & Smola, AJ 2009, Robust near-isometric matching via structured learning of graphical models. in Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference. Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference, Neural Information Processing Systems, pp. 1057-1064, 22nd Annual Conference on Neural Information Processing Systems, NIPS 2008, Vancouver, BC, Canada, 8/12/08.

Robust near-isometric matching via structured learning of graphical models. / McAuley, Julian J.; Caetano, Tibério S.; Smola, Alexander J.
Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference. Neural Information Processing Systems, 2009. p. 1057-1064 (Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Robust near-isometric matching via structured learning of graphical models

AU - McAuley, Julian J.

AU - Caetano, Tibério S.

AU - Smola, Alexander J.

PY - 2009

Y1 - 2009

N2 - Models for near-rigid shape matching are typically based on distance-related features, in order to infer matches that are consistent with the isometric assumption. However, real shapes from image datasets, even when expected to be related by "almost isometric" transformations, are actually subject not only to noise but also, to some limited degree, to variations in appearance and scale. In this paper, we introduce a graphical model that parameterises appearance, distance, and angle features and we learn all of the involved parameters via structured prediction. The outcome is a model for near-rigid shape matching which is robust in the sense that it is able to capture the possibly limited but still important scale and appearance variations. Our experimental results reveal substantial improvements upon recent successful models, while maintaining similar running times.

AB - Models for near-rigid shape matching are typically based on distance-related features, in order to infer matches that are consistent with the isometric assumption. However, real shapes from image datasets, even when expected to be related by "almost isometric" transformations, are actually subject not only to noise but also, to some limited degree, to variations in appearance and scale. In this paper, we introduce a graphical model that parameterises appearance, distance, and angle features and we learn all of the involved parameters via structured prediction. The outcome is a model for near-rigid shape matching which is robust in the sense that it is able to capture the possibly limited but still important scale and appearance variations. Our experimental results reveal substantial improvements upon recent successful models, while maintaining similar running times.

UR - http://www.scopus.com/inward/record.url?scp=84858791109&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781605609492

T3 - Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference

SP - 1057

EP - 1064

BT - Advances in Neural Information Processing Systems 21 - Proceedings of the 2008 Conference

PB - Neural Information Processing Systems

T2 - 22nd Annual Conference on Neural Information Processing Systems, NIPS 2008

Y2 - 8 December 2008 through 11 December 2008

ER -

Robust near-isometric matching via structured learning of graphical models

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this