A constrained probabilistic Petri Net framework for human activity detection in video

Massimiliano Albanese, Rama Chellappa, Vincenzo Moscato, Antonio Picariello, V. S. Subrahmanian, Pavan Turaga, Octavian Udrea

Research output: Contribution to journalArticle

46 Citations (Scopus)

Abstract

Recognition of human activities in restricted settings such as airports, parking lots and banks is of significant interest in security and automated surveillance systems. In such settings, data is usually in the form of surveillance videos with wide variation in quality and granularity. Interpretation and identification of human activities requires an activity model that a) is rich enough to handle complex multi-agent interactions, b) is robust to uncertainty in low-level processing and c) can handle ambiguities in the unfolding of activities. We present a computational framework for human activity representation based on Petri nets. We propose an extensionProbabilistic Petri Nets (PPN)and show how this model is well suited to address each of the above requirements in a wide variety of settings. We then focus on answering two types of questions: (i) what are the minimal sub-videos in which a given activity is identified with a probability above a certain threshold and (ii) for a given video, which activity from a given set occurred with the highest probability? We provide the PPN-MPS algorithm for the first problem, as well as two different algorithms (naive PPN-MPA and PPN-MPA) to solve the second. Our experimental results on a dataset consisting of bank surveillance videos and an unconstrained TSA tarmac surveillance dataset show that our algorithms are both fast and provide high quality results.

Original languageEnglish (US)
Article number4694845
Pages (from-to)1429-1443
Number of pages15
JournalIEEE Transactions on Multimedia
Volume10
Issue number8
DOIs
StatePublished - Dec 2008
Externally publishedYes

Fingerprint

Petri nets
Parking
Airports
Identification (control systems)
Processing

Keywords

  • Algorithms
  • Machine vision
  • Petri nets
  • Surveillance

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing
  • Media Technology
  • Computer Science Applications

Cite this

Albanese, M., Chellappa, R., Moscato, V., Picariello, A., Subrahmanian, V. S., Turaga, P., & Udrea, O. (2008). A constrained probabilistic Petri Net framework for human activity detection in video. IEEE Transactions on Multimedia, 10(8), 1429-1443. [4694845]. https://doi.org/10.1109/TMM.2008.2010417

A constrained probabilistic Petri Net framework for human activity detection in video. / Albanese, Massimiliano; Chellappa, Rama; Moscato, Vincenzo; Picariello, Antonio; Subrahmanian, V. S.; Turaga, Pavan; Udrea, Octavian.

In: IEEE Transactions on Multimedia, Vol. 10, No. 8, 4694845, 12.2008, p. 1429-1443.

Research output: Contribution to journalArticle

Albanese, M, Chellappa, R, Moscato, V, Picariello, A, Subrahmanian, VS, Turaga, P & Udrea, O 2008, 'A constrained probabilistic Petri Net framework for human activity detection in video', IEEE Transactions on Multimedia, vol. 10, no. 8, 4694845, pp. 1429-1443. https://doi.org/10.1109/TMM.2008.2010417
Albanese M, Chellappa R, Moscato V, Picariello A, Subrahmanian VS, Turaga P et al. A constrained probabilistic Petri Net framework for human activity detection in video. IEEE Transactions on Multimedia. 2008 Dec;10(8):1429-1443. 4694845. https://doi.org/10.1109/TMM.2008.2010417
Albanese, Massimiliano ; Chellappa, Rama ; Moscato, Vincenzo ; Picariello, Antonio ; Subrahmanian, V. S. ; Turaga, Pavan ; Udrea, Octavian. / A constrained probabilistic Petri Net framework for human activity detection in video. In: IEEE Transactions on Multimedia. 2008 ; Vol. 10, No. 8. pp. 1429-1443.
@article{adc16280e4724afa877cd16acdd5e2d0,
title = "A constrained probabilistic Petri Net framework for human activity detection in video",
abstract = "Recognition of human activities in restricted settings such as airports, parking lots and banks is of significant interest in security and automated surveillance systems. In such settings, data is usually in the form of surveillance videos with wide variation in quality and granularity. Interpretation and identification of human activities requires an activity model that a) is rich enough to handle complex multi-agent interactions, b) is robust to uncertainty in low-level processing and c) can handle ambiguities in the unfolding of activities. We present a computational framework for human activity representation based on Petri nets. We propose an extensionProbabilistic Petri Nets (PPN)and show how this model is well suited to address each of the above requirements in a wide variety of settings. We then focus on answering two types of questions: (i) what are the minimal sub-videos in which a given activity is identified with a probability above a certain threshold and (ii) for a given video, which activity from a given set occurred with the highest probability? We provide the PPN-MPS algorithm for the first problem, as well as two different algorithms (naive PPN-MPA and PPN-MPA) to solve the second. Our experimental results on a dataset consisting of bank surveillance videos and an unconstrained TSA tarmac surveillance dataset show that our algorithms are both fast and provide high quality results.",
keywords = "Algorithms, Machine vision, Petri nets, Surveillance",
author = "Massimiliano Albanese and Rama Chellappa and Vincenzo Moscato and Antonio Picariello and Subrahmanian, {V. S.} and Pavan Turaga and Octavian Udrea",
year = "2008",
month = "12",
doi = "10.1109/TMM.2008.2010417",
language = "English (US)",
volume = "10",
pages = "1429--1443",
journal = "IEEE Transactions on Multimedia",
issn = "1520-9210",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "8",

}

TY - JOUR

T1 - A constrained probabilistic Petri Net framework for human activity detection in video

AU - Albanese, Massimiliano

AU - Chellappa, Rama

AU - Moscato, Vincenzo

AU - Picariello, Antonio

AU - Subrahmanian, V. S.

AU - Turaga, Pavan

AU - Udrea, Octavian

PY - 2008/12

Y1 - 2008/12

N2 - Recognition of human activities in restricted settings such as airports, parking lots and banks is of significant interest in security and automated surveillance systems. In such settings, data is usually in the form of surveillance videos with wide variation in quality and granularity. Interpretation and identification of human activities requires an activity model that a) is rich enough to handle complex multi-agent interactions, b) is robust to uncertainty in low-level processing and c) can handle ambiguities in the unfolding of activities. We present a computational framework for human activity representation based on Petri nets. We propose an extensionProbabilistic Petri Nets (PPN)and show how this model is well suited to address each of the above requirements in a wide variety of settings. We then focus on answering two types of questions: (i) what are the minimal sub-videos in which a given activity is identified with a probability above a certain threshold and (ii) for a given video, which activity from a given set occurred with the highest probability? We provide the PPN-MPS algorithm for the first problem, as well as two different algorithms (naive PPN-MPA and PPN-MPA) to solve the second. Our experimental results on a dataset consisting of bank surveillance videos and an unconstrained TSA tarmac surveillance dataset show that our algorithms are both fast and provide high quality results.

AB - Recognition of human activities in restricted settings such as airports, parking lots and banks is of significant interest in security and automated surveillance systems. In such settings, data is usually in the form of surveillance videos with wide variation in quality and granularity. Interpretation and identification of human activities requires an activity model that a) is rich enough to handle complex multi-agent interactions, b) is robust to uncertainty in low-level processing and c) can handle ambiguities in the unfolding of activities. We present a computational framework for human activity representation based on Petri nets. We propose an extensionProbabilistic Petri Nets (PPN)and show how this model is well suited to address each of the above requirements in a wide variety of settings. We then focus on answering two types of questions: (i) what are the minimal sub-videos in which a given activity is identified with a probability above a certain threshold and (ii) for a given video, which activity from a given set occurred with the highest probability? We provide the PPN-MPS algorithm for the first problem, as well as two different algorithms (naive PPN-MPA and PPN-MPA) to solve the second. Our experimental results on a dataset consisting of bank surveillance videos and an unconstrained TSA tarmac surveillance dataset show that our algorithms are both fast and provide high quality results.

KW - Algorithms

KW - Machine vision

KW - Petri nets

KW - Surveillance

UR - http://www.scopus.com/inward/record.url?scp=57849120833&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=57849120833&partnerID=8YFLogxK

U2 - 10.1109/TMM.2008.2010417

DO - 10.1109/TMM.2008.2010417

M3 - Article

AN - SCOPUS:57849120833

VL - 10

SP - 1429

EP - 1443

JO - IEEE Transactions on Multimedia

JF - IEEE Transactions on Multimedia

SN - 1520-9210

IS - 8

M1 - 4694845

ER -