Full-Angle Quaternions for Robustly Matching Vectors of 3D Rotations

Stephan Liwicki; Minh-Tri Pham; Stefanos Zafeiriou; Maja Pantic; Björn Stenger

doi:10.1109/CVPR.2014.21

Full-Angle Quaternions for Robustly Matching Vectors of 3D Rotations

Stephan Liwicki
, Minh-Tri Pham
, Stefanos Zafeiriou
, Maja Pantic
, Björn Stenger

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

3 Citations (Scopus)

258 Downloads (Pure)

Abstract

In this paper we introduce a new distance for robustly matching vectors of 3D rotations. A special representation of 3D rotations, which we coin full-angle quaternion (FAQ), allows us to express this distance as Euclidean. We apply the distance to the problems of 3D shape recognition from point clouds and 2D object tracking in color video. For the former, we introduce a hashing scheme for scale and translation which outperforms the previous state-of-the-art approach on a public dataset. For the latter, we incorporate online subspace learning with the proposed FAQ representation to highlight the benefits of the new representation.

Original language	Undefined
Title of host publication	Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014
Place of Publication	USA
Publisher	IEEE
Pages	105-112
Number of pages	8
ISBN (Print)	978-1-4799-5117-8
DOIs	https://doi.org/10.1109/CVPR.2014.21
Publication status	Published - Jun 2014
Event	27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014 - Columbus, OH, USA, Columbus, United States Duration: 23 Jun 2014 → 28 Jun 2014 Conference number: 27

Publication series

Name
Publisher	IEEE Computer Society
ISSN (Print)	1063-6919

Conference

Conference	27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014
Abbreviated title	CVPR 2014
Country/Territory	United States
City	Columbus
Period	23/06/14 → 28/06/14
Other	23-28 June 2014

Keywords

HMI-HF: Human Factors
METIS-309945
IR-95226
EWI-25819

Access to Document

10.1109/CVPR.2014.21

Pantic_Full-Angle_Quaternions

Cite this

@inproceedings{65c28812d7a040968fbe63239839d383,

title = "Full-Angle Quaternions for Robustly Matching Vectors of 3D Rotations",

abstract = "In this paper we introduce a new distance for robustly matching vectors of 3D rotations. A special representation of 3D rotations, which we coin full-angle quaternion (FAQ), allows us to express this distance as Euclidean. We apply the distance to the problems of 3D shape recognition from point clouds and 2D object tracking in color video. For the former, we introduce a hashing scheme for scale and translation which outperforms the previous state-of-the-art approach on a public dataset. For the latter, we incorporate online subspace learning with the proposed FAQ representation to highlight the benefits of the new representation.",

keywords = "HMI-HF: Human Factors, METIS-309945, IR-95226, EWI-25819",

author = "Stephan Liwicki and Minh-Tri Pham and Stefanos Zafeiriou and Maja Pantic and Bj{\"o}rn Stenger",

note = "10.1109/CVPR.2014.21 ; 27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014 ; Conference date: 23-06-2014 Through 28-06-2014",

year = "2014",

month = jun,

doi = "10.1109/CVPR.2014.21",

language = "Undefined",

isbn = "978-1-4799-5117-8",

publisher = "IEEE",

pages = "105--112",

booktitle = "Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014",

address = "United States",

}

Liwicki, S, Pham, M-T, Zafeiriou, S, Pantic, M & Stenger, B 2014, Full-Angle Quaternions for Robustly Matching Vectors of 3D Rotations. in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014. IEEE, USA, pp. 105-112, 27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014, Columbus, Ohio, United States, 23/06/14. https://doi.org/10.1109/CVPR.2014.21

TY - GEN

T1 - Full-Angle Quaternions for Robustly Matching Vectors of 3D Rotations

AU - Liwicki, Stephan

AU - Pham, Minh-Tri

AU - Zafeiriou, Stefanos

AU - Pantic, Maja

AU - Stenger, Björn

N1 - Conference code: 27

PY - 2014/6

Y1 - 2014/6

N2 - In this paper we introduce a new distance for robustly matching vectors of 3D rotations. A special representation of 3D rotations, which we coin full-angle quaternion (FAQ), allows us to express this distance as Euclidean. We apply the distance to the problems of 3D shape recognition from point clouds and 2D object tracking in color video. For the former, we introduce a hashing scheme for scale and translation which outperforms the previous state-of-the-art approach on a public dataset. For the latter, we incorporate online subspace learning with the proposed FAQ representation to highlight the benefits of the new representation.

AB - In this paper we introduce a new distance for robustly matching vectors of 3D rotations. A special representation of 3D rotations, which we coin full-angle quaternion (FAQ), allows us to express this distance as Euclidean. We apply the distance to the problems of 3D shape recognition from point clouds and 2D object tracking in color video. For the former, we introduce a hashing scheme for scale and translation which outperforms the previous state-of-the-art approach on a public dataset. For the latter, we incorporate online subspace learning with the proposed FAQ representation to highlight the benefits of the new representation.

KW - HMI-HF: Human Factors

KW - METIS-309945

KW - IR-95226

KW - EWI-25819

U2 - 10.1109/CVPR.2014.21

DO - 10.1109/CVPR.2014.21

M3 - Conference contribution

SN - 978-1-4799-5117-8

SP - 105

EP - 112

BT - Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014

PB - IEEE

CY - USA

T2 - 27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014

Y2 - 23 June 2014 through 28 June 2014

ER -