Image classification using super-vector coding of local image descriptors

Xi Zhou; Kai Yu; Tong Zhang; Thomas S. Huang

doi:10.1007/978-3-642-15555-0_11

Image classification using super-vector coding of local image descriptors

Xi Zhou, Kai Yu, Tong Zhang, Thomas S. Huang

Coordinated Science Lab

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

Abstract

This paper introduces a new framework for image classification using local visual descriptors. The pipeline first performs a nonlinear feature transformation on descriptors, then aggregates the results together to form image-level representations, and finally applies a classification model. For all the three steps we suggest novel solutions which make our approach appealing in theory, more scalable in computation, and transparent in classification. Our experiments demonstrate that the proposed classification method achieves state-of-the-art accuracy on the well-known PASCAL benchmarks.

Original language	English (US)
Title of host publication	Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings
Publisher	Springer
Pages	141-154
Number of pages	14
Edition	PART 5
ISBN (Print)	3642155545, 9783642155543
DOIs	https://doi.org/10.1007/978-3-642-15555-0_11
State	Published - 2010
Event	11th European Conference on Computer Vision, ECCV 2010 - Heraklion, Crete, Greece Duration: Sep 10 2010 → Sep 11 2010

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 5
Volume	6315 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th European Conference on Computer Vision, ECCV 2010
Country/Territory	Greece
City	Heraklion, Crete
Period	9/10/10 → 9/11/10

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Online availability

10.1007/978-3-642-15555-0_11

Library availability

Discover UIUC Full Text

Cite this

Zhou, X., Yu, K., Zhang, T., & Huang, T. S. (2010). Image classification using super-vector coding of local image descriptors. In Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings (PART 5 ed., pp. 141-154). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6315 LNCS, No. PART 5). Springer. https://doi.org/10.1007/978-3-642-15555-0_11

Image classification using super-vector coding of local image descriptors. / Zhou, Xi; Yu, Kai; Zhang, Tong et al.
Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings. PART 5. ed. Springer, 2010. p. 141-154 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6315 LNCS, No. PART 5).

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

Zhou, X, Yu, K, Zhang, T & Huang, TS 2010, Image classification using super-vector coding of local image descriptors. in Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings. PART 5 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 5, vol. 6315 LNCS, Springer, pp. 141-154, 11th European Conference on Computer Vision, ECCV 2010, Heraklion, Crete, Greece, 9/10/10. https://doi.org/10.1007/978-3-642-15555-0_11

Zhou X, Yu K, Zhang T, Huang TS. Image classification using super-vector coding of local image descriptors. In Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings. PART 5 ed. Springer. 2010. p. 141-154. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 5). doi: 10.1007/978-3-642-15555-0_11

Zhou, Xi ; Yu, Kai ; Zhang, Tong et al. / Image classification using super-vector coding of local image descriptors. Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings. PART 5. ed. Springer, 2010. pp. 141-154 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 5).

@inproceedings{a252b8ac0bb84f73ad403842be289ed7,

title = "Image classification using super-vector coding of local image descriptors",

abstract = "This paper introduces a new framework for image classification using local visual descriptors. The pipeline first performs a nonlinear feature transformation on descriptors, then aggregates the results together to form image-level representations, and finally applies a classification model. For all the three steps we suggest novel solutions which make our approach appealing in theory, more scalable in computation, and transparent in classification. Our experiments demonstrate that the proposed classification method achieves state-of-the-art accuracy on the well-known PASCAL benchmarks.",

author = "Xi Zhou and Kai Yu and Tong Zhang and Huang, {Thomas S.}",

year = "2010",

doi = "10.1007/978-3-642-15555-0_11",

language = "English (US)",

isbn = "3642155545",

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

publisher = "Springer",

number = "PART 5",

pages = "141--154",

booktitle = "Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings",

address = "Germany",

edition = "PART 5",

}

TY - GEN

T1 - Image classification using super-vector coding of local image descriptors

AU - Zhou, Xi

AU - Yu, Kai

AU - Zhang, Tong

AU - Huang, Thomas S.

PY - 2010

Y1 - 2010

N2 - This paper introduces a new framework for image classification using local visual descriptors. The pipeline first performs a nonlinear feature transformation on descriptors, then aggregates the results together to form image-level representations, and finally applies a classification model. For all the three steps we suggest novel solutions which make our approach appealing in theory, more scalable in computation, and transparent in classification. Our experiments demonstrate that the proposed classification method achieves state-of-the-art accuracy on the well-known PASCAL benchmarks.

AB - This paper introduces a new framework for image classification using local visual descriptors. The pipeline first performs a nonlinear feature transformation on descriptors, then aggregates the results together to form image-level representations, and finally applies a classification model. For all the three steps we suggest novel solutions which make our approach appealing in theory, more scalable in computation, and transparent in classification. Our experiments demonstrate that the proposed classification method achieves state-of-the-art accuracy on the well-known PASCAL benchmarks.

UR - http://www.scopus.com/inward/record.url?scp=78149344615&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78149344615&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-15555-0_11

DO - 10.1007/978-3-642-15555-0_11

M3 - Conference contribution

AN - SCOPUS:78149344615

SN - 3642155545

SN - 9783642155543

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

SP - 141

EP - 154

BT - Computer Vision, ECCV 2010 - 11th European Conference on Computer Vision, Proceedings

PB - Springer

T2 - 11th European Conference on Computer Vision, ECCV 2010

Y2 - 10 September 2010 through 11 September 2010

ER -

Image classification using super-vector coding of local image descriptors

Abstract

Publication series

Conference

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this