Simultaneous tracking and activity recognition

Cristina Elena Manfredotti; David J. Fleet; Howard J. Hamilton; Sandra Zilles

doi:10.1109/ICTAI.2011.36

Simultaneous tracking and activity recognition

Cristina Elena Manfredotti, David J. Fleet, Howard J. Hamilton, Sandra Zilles

Department of Computer Science

4 Citations (Scopus)

Abstract

Many tracking problems involve several distinct objects interacting with each other. We develop a framework that takes into account interactions between objects allowing the recognition of complex activities. In contrast to classic approaches that consider distinct phases of tracking and activity recognition, our framework performs these two tasks simultaneously. In particular, we adopt a Bayesian standpoint where the system maintains a joint distribution of the positions, the interactions and the possible activities. This turns out to be advantegeous, as information about the ongoing activities can be used to improve the prediction step of the tracking, while, at the same time, tracking information can be used for online activity recognition. Experimental results in two different settings show that our approach 1) decreases the error rate and improves the identity maintenance of the positional tracking and 2) identifies the correct activity with higher accuracy than standard approaches.

Original language	English
Title of host publication	Proceedings - 2011 23rd IEEE International Conference on Tools with Artificial Intelligence : ICTAI 2011
Number of pages	8
Publisher	IEEE
Publication date	2011
Pages	189-196
ISBN (Print)	978-0-7695-4596-7
DOIs	https://doi.org/10.1109/ICTAI.2011.36
Publication status	Published - 2011
Event	23rd IEEE International Conference on Tools with Artificial Intelligence - Boca Raton, United States Duration: 7 Nov 2011 → 9 Nov 2011 Conference number: 23

Conference

Conference	23rd IEEE International Conference on Tools with Artificial Intelligence
Number	23
Country/Territory	United States
City	Boca Raton
Period	07/11/2011 → 09/11/2011

Series	Proceedings - International Conference on Tools with Artificial Intelligence, TAI
ISSN	1082-3409

Access to Document

10.1109/ICTAI.2011.36

Cite this

Simultaneous tracking and activity recognition. / Manfredotti, Cristina Elena; Fleet, David J.; Hamilton, Howard J. et al.

Proceedings - 2011 23rd IEEE International Conference on Tools with Artificial Intelligence: ICTAI 2011. IEEE, 2011. p. 189-196 (Proceedings - International Conference on Tools with Artificial Intelligence, TAI ).

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

Manfredotti, CE, Fleet, DJ, Hamilton, HJ & Zilles, S 2011, Simultaneous tracking and activity recognition. in Proceedings - 2011 23rd IEEE International Conference on Tools with Artificial Intelligence: ICTAI 2011. IEEE, Proceedings - International Conference on Tools with Artificial Intelligence, TAI , pp. 189-196, 23rd IEEE International Conference on Tools with Artificial Intelligence, Boca Raton, United States, 07/11/2011. https://doi.org/10.1109/ICTAI.2011.36

@inproceedings{6b6ef4ce0aff4d53a0e2985a4c9cb93d,

title = "Simultaneous tracking and activity recognition",

abstract = "Many tracking problems involve several distinct objects interacting with each other. We develop a framework that takes into account interactions between objects allowing the recognition of complex activities. In contrast to classic approaches that consider distinct phases of tracking and activity recognition, our framework performs these two tasks simultaneously. In particular, we adopt a Bayesian standpoint where the system maintains a joint distribution of the positions, the interactions and the possible activities. This turns out to be advantegeous, as information about the ongoing activities can be used to improve the prediction step of the tracking, while, at the same time, tracking information can be used for online activity recognition. Experimental results in two different settings show that our approach 1) decreases the error rate and improves the identity maintenance of the positional tracking and 2) identifies the correct activity with higher accuracy than standard approaches.",

author = "Manfredotti, {Cristina Elena} and Fleet, {David J.} and Hamilton, {Howard J.} and Sandra Zilles",

year = "2011",

doi = "10.1109/ICTAI.2011.36",

language = "English",

isbn = "978-0-7695-4596-7",

series = "Proceedings - International Conference on Tools with Artificial Intelligence, TAI ",

publisher = "IEEE",

pages = "189--196",

booktitle = "Proceedings - 2011 23rd IEEE International Conference on Tools with Artificial Intelligence",

note = "23rd IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2011 ; Conference date: 07-11-2011 Through 09-11-2011",

}

TY - GEN

T1 - Simultaneous tracking and activity recognition

AU - Manfredotti, Cristina Elena

AU - Fleet, David J.

AU - Hamilton, Howard J.

AU - Zilles, Sandra

N1 - Conference code: 23

PY - 2011

Y1 - 2011

N2 - Many tracking problems involve several distinct objects interacting with each other. We develop a framework that takes into account interactions between objects allowing the recognition of complex activities. In contrast to classic approaches that consider distinct phases of tracking and activity recognition, our framework performs these two tasks simultaneously. In particular, we adopt a Bayesian standpoint where the system maintains a joint distribution of the positions, the interactions and the possible activities. This turns out to be advantegeous, as information about the ongoing activities can be used to improve the prediction step of the tracking, while, at the same time, tracking information can be used for online activity recognition. Experimental results in two different settings show that our approach 1) decreases the error rate and improves the identity maintenance of the positional tracking and 2) identifies the correct activity with higher accuracy than standard approaches.

AB - Many tracking problems involve several distinct objects interacting with each other. We develop a framework that takes into account interactions between objects allowing the recognition of complex activities. In contrast to classic approaches that consider distinct phases of tracking and activity recognition, our framework performs these two tasks simultaneously. In particular, we adopt a Bayesian standpoint where the system maintains a joint distribution of the positions, the interactions and the possible activities. This turns out to be advantegeous, as information about the ongoing activities can be used to improve the prediction step of the tracking, while, at the same time, tracking information can be used for online activity recognition. Experimental results in two different settings show that our approach 1) decreases the error rate and improves the identity maintenance of the positional tracking and 2) identifies the correct activity with higher accuracy than standard approaches.

U2 - 10.1109/ICTAI.2011.36

DO - 10.1109/ICTAI.2011.36

M3 - Article in proceedings

SN - 978-0-7695-4596-7

T3 - Proceedings - International Conference on Tools with Artificial Intelligence, TAI

SP - 189

EP - 196

BT - Proceedings - 2011 23rd IEEE International Conference on Tools with Artificial Intelligence

PB - IEEE

T2 - 23rd IEEE International Conference on Tools with Artificial Intelligence

Y2 - 7 November 2011 through 9 November 2011

ER -