Spectral classification of 3D articulated shapes

Zhenbao Liu; Feng Zhang; Shuhui Bu

doi:10.1007/978-3-319-04117-9_30

Spectral classification of 3D articulated shapes

Zhenbao Liu, Feng Zhang, Shuhui Bu

School of Civil Aviation

Northwestern Polytechnical University Xian

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

1 Scopus citations

Abstract

A large number of 3D models distributed on internet has created the demand for automatic shape classification. This paper presents a novel classification method for 3D mesh shapes. Each shape is represented by the eigenvalues of an appropriately defined affinity matrix, forming a spectral embedding which achieves invariance against rigid-body transformations, uniform scaling, and shape articulation. And then, Adaboost algorithm is applied to classify the 3D models in the spectral space according to its immunity to overfitting. We evaluate the approach on the McGill 3D shape benchmark and compare the results with previous classification method, and it achieves higher classification accuracy. This method is suitable for automatic classification of 3D articulated shapes.

Original language	English
Title of host publication	MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings
Pages	315-322
Number of pages	8
Edition	PART 2
DOIs	https://doi.org/10.1007/978-3-319-04117-9_30
State	Published - 2014
Event	20th Anniversary International Conference on MultiMedia Modeling, MMM 2014 - Dublin, Ireland Duration: 6 Jan 2014 → 10 Jan 2014

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	8326 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th Anniversary International Conference on MultiMedia Modeling, MMM 2014
Country/Territory	Ireland
City	Dublin
Period	6/01/14 → 10/01/14

Keywords

3D Shape
Boosting
Spectral classification

Access to Document

10.1007/978-3-319-04117-9_30

Cite this

Liu, Z., Zhang, F., & Bu, S. (2014). Spectral classification of 3D articulated shapes. In MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings (PART 2 ed., pp. 315-322). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8326 LNCS, No. PART 2). https://doi.org/10.1007/978-3-319-04117-9_30

@inproceedings{5306f0278fef4788adf452d01c1e27c2,

title = "Spectral classification of 3D articulated shapes",

abstract = "A large number of 3D models distributed on internet has created the demand for automatic shape classification. This paper presents a novel classification method for 3D mesh shapes. Each shape is represented by the eigenvalues of an appropriately defined affinity matrix, forming a spectral embedding which achieves invariance against rigid-body transformations, uniform scaling, and shape articulation. And then, Adaboost algorithm is applied to classify the 3D models in the spectral space according to its immunity to overfitting. We evaluate the approach on the McGill 3D shape benchmark and compare the results with previous classification method, and it achieves higher classification accuracy. This method is suitable for automatic classification of 3D articulated shapes.",

keywords = "3D Shape, Boosting, Spectral classification",

author = "Zhenbao Liu and Feng Zhang and Shuhui Bu",

year = "2014",

doi = "10.1007/978-3-319-04117-9_30",

language = "英语",

isbn = "9783319041162",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 2",

pages = "315--322",

booktitle = "MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings",

edition = "PART 2",

note = "20th Anniversary International Conference on MultiMedia Modeling, MMM 2014 ; Conference date: 06-01-2014 Through 10-01-2014",

}

Liu, Z, Zhang, F & Bu, S 2014, Spectral classification of 3D articulated shapes. in MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 8326 LNCS, pp. 315-322, 20th Anniversary International Conference on MultiMedia Modeling, MMM 2014, Dublin, Ireland, 6/01/14. https://doi.org/10.1007/978-3-319-04117-9_30

Spectral classification of 3D articulated shapes. / Liu, Zhenbao; Zhang, Feng; Bu, Shuhui.
MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings. PART 2. ed. 2014. p. 315-322 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8326 LNCS, No. PART 2).

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

TY - GEN

T1 - Spectral classification of 3D articulated shapes

AU - Liu, Zhenbao

AU - Zhang, Feng

AU - Bu, Shuhui

PY - 2014

Y1 - 2014

N2 - A large number of 3D models distributed on internet has created the demand for automatic shape classification. This paper presents a novel classification method for 3D mesh shapes. Each shape is represented by the eigenvalues of an appropriately defined affinity matrix, forming a spectral embedding which achieves invariance against rigid-body transformations, uniform scaling, and shape articulation. And then, Adaboost algorithm is applied to classify the 3D models in the spectral space according to its immunity to overfitting. We evaluate the approach on the McGill 3D shape benchmark and compare the results with previous classification method, and it achieves higher classification accuracy. This method is suitable for automatic classification of 3D articulated shapes.

AB - A large number of 3D models distributed on internet has created the demand for automatic shape classification. This paper presents a novel classification method for 3D mesh shapes. Each shape is represented by the eigenvalues of an appropriately defined affinity matrix, forming a spectral embedding which achieves invariance against rigid-body transformations, uniform scaling, and shape articulation. And then, Adaboost algorithm is applied to classify the 3D models in the spectral space according to its immunity to overfitting. We evaluate the approach on the McGill 3D shape benchmark and compare the results with previous classification method, and it achieves higher classification accuracy. This method is suitable for automatic classification of 3D articulated shapes.

KW - 3D Shape

KW - Boosting

KW - Spectral classification

UR - http://www.scopus.com/inward/record.url?scp=84893474374&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-04117-9_30

DO - 10.1007/978-3-319-04117-9_30

M3 - 会议稿件

AN - SCOPUS:84893474374

SN - 9783319041162

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 315

EP - 322

BT - MultiMedia Modeling - 20th Anniversary International Conference, MMM 2014, Proceedings

T2 - 20th Anniversary International Conference on MultiMedia Modeling, MMM 2014

Y2 - 6 January 2014 through 10 January 2014

ER -

Spectral classification of 3D articulated shapes

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this