A piggyback representation for action recognition

Lior Wolf; Yair Hanani; Tal Hassner

doi:10.1109/CVPRW.2014.81

A piggyback representation for action recognition

Lior Wolf^*, Yair Hanani, Tal Hassner

^*Corresponding author for this work

School of Computer Science

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

2 Scopus citations

Abstract

In video understanding, the spatial patterns formed by local space-time interest points hold discriminative information. We encode these spatial regularities using a word2vec neural network, a recently proposed tool in the field of text processing. Then, building upon recent accumulator based image representation solutions, input videos are represented in a hybrid manner: the appearance of local space time interest points is used to collect and associate the learned descriptors, which capture the spatial patterns. Promising results are shown on recent action recognition benchmarks, using well established methods as the underlying appearance descriptors.

Original language	English
Title of host publication	Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014
Publisher	IEEE Computer Society
Pages	520-525
Number of pages	6
ISBN (Electronic)	9781479943098, 9781479943098
DOIs	https://doi.org/10.1109/CVPRW.2014.81
State	Published - 24 Sep 2014
Event	2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014 - Columbus, United States Duration: 23 Jun 2014 → 28 Jun 2014

Publication series

Name	IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
ISSN (Print)	2160-7508
ISSN (Electronic)	2160-7516

Conference

Conference	2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014
Country/Territory	United States
City	Columbus
Period	23/06/14 → 28/06/14

Access to Document

10.1109/CVPRW.2014.81

Cite this

Wolf, L., Hanani, Y., & Hassner, T. (2014). A piggyback representation for action recognition. In Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014 (pp. 520-525). Article 6910030 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops). IEEE Computer Society. https://doi.org/10.1109/CVPRW.2014.81

@inproceedings{7629fc7f5bc14d98bed02d7aa72dba36,

title = "A piggyback representation for action recognition",

abstract = "In video understanding, the spatial patterns formed by local space-time interest points hold discriminative information. We encode these spatial regularities using a word2vec neural network, a recently proposed tool in the field of text processing. Then, building upon recent accumulator based image representation solutions, input videos are represented in a hybrid manner: the appearance of local space time interest points is used to collect and associate the learned descriptors, which capture the spatial patterns. Promising results are shown on recent action recognition benchmarks, using well established methods as the underlying appearance descriptors.",

author = "Lior Wolf and Yair Hanani and Tal Hassner",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014 ; Conference date: 23-06-2014 Through 28-06-2014",

year = "2014",

month = sep,

day = "24",

doi = "10.1109/CVPRW.2014.81",

language = "אנגלית",

series = "IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops",

publisher = "IEEE Computer Society",

pages = "520--525",

booktitle = "Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014",

address = "ארצות הברית",

}

Wolf, L, Hanani, Y & Hassner, T 2014, A piggyback representation for action recognition. in Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014., 6910030, IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, IEEE Computer Society, pp. 520-525, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014, Columbus, United States, 23/06/14. https://doi.org/10.1109/CVPRW.2014.81

A piggyback representation for action recognition. / Wolf, Lior; Hanani, Yair; Hassner, Tal.
Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014. IEEE Computer Society, 2014. p. 520-525 6910030 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops).

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

TY - GEN

T1 - A piggyback representation for action recognition

AU - Wolf, Lior

AU - Hanani, Yair

AU - Hassner, Tal

PY - 2014/9/24

Y1 - 2014/9/24

N2 - In video understanding, the spatial patterns formed by local space-time interest points hold discriminative information. We encode these spatial regularities using a word2vec neural network, a recently proposed tool in the field of text processing. Then, building upon recent accumulator based image representation solutions, input videos are represented in a hybrid manner: the appearance of local space time interest points is used to collect and associate the learned descriptors, which capture the spatial patterns. Promising results are shown on recent action recognition benchmarks, using well established methods as the underlying appearance descriptors.

AB - In video understanding, the spatial patterns formed by local space-time interest points hold discriminative information. We encode these spatial regularities using a word2vec neural network, a recently proposed tool in the field of text processing. Then, building upon recent accumulator based image representation solutions, input videos are represented in a hybrid manner: the appearance of local space time interest points is used to collect and associate the learned descriptors, which capture the spatial patterns. Promising results are shown on recent action recognition benchmarks, using well established methods as the underlying appearance descriptors.

UR - http://www.scopus.com/inward/record.url?scp=84908520907&partnerID=8YFLogxK

U2 - 10.1109/CVPRW.2014.81

DO - 10.1109/CVPRW.2014.81

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:84908520907

T3 - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops

SP - 520

EP - 525

BT - Proceedings - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014

PB - IEEE Computer Society

T2 - 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2014

Y2 - 23 June 2014 through 28 June 2014

ER -

A piggyback representation for action recognition

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this