On the sensitivity of spatio-temporal interest points to person identity

Mouna Selmi; Mounim El Yacoubi; Bernadette Dorizzi

doi:10.1109/SSIAI.2012.6202455

On the sensitivity of spatio-temporal interest points to person identity

Mouna Selmi
, Mounim El Yacoubi
, Bernadette Dorizzi

Telecom Paris

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

1 Citation (Scopus)

Abstract

Spatio-temporal interest points (STIPs) have recently become a mainstream technique for encoding human action video sequences. These local features overcome the complex issues of background subtraction and human tracking, and encode relevant motion information for recognition. Nonetheless, STIPs may result not only from actions themselves but from other factors such as person identity. This paper addresses modeling the multi-modal aspect of motions using Multilinear Tensor Analysis, with the goal of detaching parameters that are relevant to action recognition from those that are irrelevant (related to person identity). This modeling of space time interest points by a tensor representation is a new contribution to human action analysis. Experiments carried out on the Weizmann and KTH datasets give the interesting finding that STIP-based features are not only related to action types but may also be sensitive to person identity or style.

Original language	English
Title of host publication	2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings
Pages	69-72
Number of pages	4
DOIs	https://doi.org/10.1109/SSIAI.2012.6202455
Publication status	Published - 2 Jul 2012
Event	2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012 - Santa Fe, NM, United States Duration: 22 Apr 2012 → 24 Apr 2012

Publication series

Name	Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation

Conference

Conference	2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012
Country/Territory	United States
City	Santa Fe, NM
Period	22/04/12 → 24/04/12

Keywords

Spatio-temporal interest points
tensor

Access to Document

10.1109/SSIAI.2012.6202455

Cite this

Selmi, M., El Yacoubi, M., & Dorizzi, B. (2012). On the sensitivity of spatio-temporal interest points to person identity. In 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings (pp. 69-72). Article 6202455 (Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation). https://doi.org/10.1109/SSIAI.2012.6202455

@inproceedings{ffe7f64b2bfb41d2b5d3017dd0c85194,

title = "On the sensitivity of spatio-temporal interest points to person identity",

abstract = "Spatio-temporal interest points (STIPs) have recently become a mainstream technique for encoding human action video sequences. These local features overcome the complex issues of background subtraction and human tracking, and encode relevant motion information for recognition. Nonetheless, STIPs may result not only from actions themselves but from other factors such as person identity. This paper addresses modeling the multi-modal aspect of motions using Multilinear Tensor Analysis, with the goal of detaching parameters that are relevant to action recognition from those that are irrelevant (related to person identity). This modeling of space time interest points by a tensor representation is a new contribution to human action analysis. Experiments carried out on the Weizmann and KTH datasets give the interesting finding that STIP-based features are not only related to action types but may also be sensitive to person identity or style.",

keywords = "Spatio-temporal interest points, tensor",

author = "Mouna Selmi and \{El Yacoubi\}, Mounim and Bernadette Dorizzi",

year = "2012",

month = jul,

day = "2",

doi = "10.1109/SSIAI.2012.6202455",

language = "English",

isbn = "9781467318303",

series = "Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation",

pages = "69--72",

booktitle = "2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings",

note = "2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012 ; Conference date: 22-04-2012 Through 24-04-2012",

}

Selmi, M, El Yacoubi, M & Dorizzi, B 2012, On the sensitivity of spatio-temporal interest points to person identity. in 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings., 6202455, Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation, pp. 69-72, 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Santa Fe, NM, United States, 22/04/12. https://doi.org/10.1109/SSIAI.2012.6202455

On the sensitivity of spatio-temporal interest points to person identity. / Selmi, Mouna; El Yacoubi, Mounim; Dorizzi, Bernadette.
2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings. 2012. p. 69-72 6202455 (Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation).

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

TY - GEN

T1 - On the sensitivity of spatio-temporal interest points to person identity

AU - Selmi, Mouna

AU - El Yacoubi, Mounim

AU - Dorizzi, Bernadette

PY - 2012/7/2

Y1 - 2012/7/2

N2 - Spatio-temporal interest points (STIPs) have recently become a mainstream technique for encoding human action video sequences. These local features overcome the complex issues of background subtraction and human tracking, and encode relevant motion information for recognition. Nonetheless, STIPs may result not only from actions themselves but from other factors such as person identity. This paper addresses modeling the multi-modal aspect of motions using Multilinear Tensor Analysis, with the goal of detaching parameters that are relevant to action recognition from those that are irrelevant (related to person identity). This modeling of space time interest points by a tensor representation is a new contribution to human action analysis. Experiments carried out on the Weizmann and KTH datasets give the interesting finding that STIP-based features are not only related to action types but may also be sensitive to person identity or style.

AB - Spatio-temporal interest points (STIPs) have recently become a mainstream technique for encoding human action video sequences. These local features overcome the complex issues of background subtraction and human tracking, and encode relevant motion information for recognition. Nonetheless, STIPs may result not only from actions themselves but from other factors such as person identity. This paper addresses modeling the multi-modal aspect of motions using Multilinear Tensor Analysis, with the goal of detaching parameters that are relevant to action recognition from those that are irrelevant (related to person identity). This modeling of space time interest points by a tensor representation is a new contribution to human action analysis. Experiments carried out on the Weizmann and KTH datasets give the interesting finding that STIP-based features are not only related to action types but may also be sensitive to person identity or style.

KW - Spatio-temporal interest points

KW - tensor

U2 - 10.1109/SSIAI.2012.6202455

DO - 10.1109/SSIAI.2012.6202455

M3 - Conference contribution

AN - SCOPUS:84862857265

SN - 9781467318303

T3 - Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation

SP - 69

EP - 72

BT - 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Proceedings

T2 - 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012

Y2 - 22 April 2012 through 24 April 2012

ER -

On the sensitivity of spatio-temporal interest points to person identity

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this