Knowledge discovery in group activities through sequential observation analysis

Vinayak Elangovan, Amir Shirkhodaie

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Understanding of Group Activities (GA) has significant applications in civilian and military domains. The process of understanding GA is typically involved with spatiotemporal analysis of multi-modality sensor data. Video imagery is one popular sensing modality that offers rich data, however, data associated with imagery source may become fragmented and discontinued due to a number of reasons (e.g., data transmission, or observation obstructions and occlusions). However, making sense out of video imagery is a real challenge. It requires a proper inference working model capable of analyzing video imagery frame by frame, extract and inference spatiotemporal information pertaining to observations while developing an incremental perception of the GA as they emerge overtime. In this paper, we propose an ontology based GA recognition where three inference Hidden Markov Models (HMM's) are used for predicting group activities taking place in outdoor environments and different task operational taxonomy. The three competing models include: a concatenated HMM, a cascaded HMM, and a context-based HMM. The proposed ontology based GA-HMM was subjected to set of semantically annotated visual observations from outdoor group activity experiments. Experimental results from GA-HMM are presented with technical discussions on design of each model and their potential implication to Persistent Surveillance Systems (PSS).

Original languageEnglish (US)
Title of host publicationSignal Processing, Sensor/Information Fusion, and Target Recognition XXIII
PublisherSPIE
ISBN (Print)9781628410280
DOIs
StatePublished - Jan 1 2014
EventSignal Processing, Sensor/Information Fusion, and Target Recognition XXIII - Baltimore, MD, United States
Duration: May 5 2014May 8 2014

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume9091
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Other

OtherSignal Processing, Sensor/Information Fusion, and Target Recognition XXIII
CountryUnited States
CityBaltimore, MD
Period5/5/145/8/14

Fingerprint

data mining
Knowledge Discovery
Hidden Markov models
Data mining
Markov Model
imagery
inference
Ontology
Taxonomies
Data communication systems
Activity Recognition
Multimodality
taxonomy
Observation
visual observation
Taxonomy
Data Transmission
occlusion
Obstruction
Occlusion

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Elangovan, V., & Shirkhodaie, A. (2014). Knowledge discovery in group activities through sequential observation analysis. In Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII [90910T] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 9091). SPIE. https://doi.org/10.1117/12.2050909
Elangovan, Vinayak ; Shirkhodaie, Amir. / Knowledge discovery in group activities through sequential observation analysis. Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII. SPIE, 2014. (Proceedings of SPIE - The International Society for Optical Engineering).
@inproceedings{55f0f747adb04aec867b214b242ede8c,
title = "Knowledge discovery in group activities through sequential observation analysis",
abstract = "Understanding of Group Activities (GA) has significant applications in civilian and military domains. The process of understanding GA is typically involved with spatiotemporal analysis of multi-modality sensor data. Video imagery is one popular sensing modality that offers rich data, however, data associated with imagery source may become fragmented and discontinued due to a number of reasons (e.g., data transmission, or observation obstructions and occlusions). However, making sense out of video imagery is a real challenge. It requires a proper inference working model capable of analyzing video imagery frame by frame, extract and inference spatiotemporal information pertaining to observations while developing an incremental perception of the GA as they emerge overtime. In this paper, we propose an ontology based GA recognition where three inference Hidden Markov Models (HMM's) are used for predicting group activities taking place in outdoor environments and different task operational taxonomy. The three competing models include: a concatenated HMM, a cascaded HMM, and a context-based HMM. The proposed ontology based GA-HMM was subjected to set of semantically annotated visual observations from outdoor group activity experiments. Experimental results from GA-HMM are presented with technical discussions on design of each model and their potential implication to Persistent Surveillance Systems (PSS).",
author = "Vinayak Elangovan and Amir Shirkhodaie",
year = "2014",
month = "1",
day = "1",
doi = "10.1117/12.2050909",
language = "English (US)",
isbn = "9781628410280",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "SPIE",
booktitle = "Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII",
address = "United States",

}

Elangovan, V & Shirkhodaie, A 2014, Knowledge discovery in group activities through sequential observation analysis. in Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII., 90910T, Proceedings of SPIE - The International Society for Optical Engineering, vol. 9091, SPIE, Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII, Baltimore, MD, United States, 5/5/14. https://doi.org/10.1117/12.2050909

Knowledge discovery in group activities through sequential observation analysis. / Elangovan, Vinayak; Shirkhodaie, Amir.

Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII. SPIE, 2014. 90910T (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 9091).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Knowledge discovery in group activities through sequential observation analysis

AU - Elangovan, Vinayak

AU - Shirkhodaie, Amir

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Understanding of Group Activities (GA) has significant applications in civilian and military domains. The process of understanding GA is typically involved with spatiotemporal analysis of multi-modality sensor data. Video imagery is one popular sensing modality that offers rich data, however, data associated with imagery source may become fragmented and discontinued due to a number of reasons (e.g., data transmission, or observation obstructions and occlusions). However, making sense out of video imagery is a real challenge. It requires a proper inference working model capable of analyzing video imagery frame by frame, extract and inference spatiotemporal information pertaining to observations while developing an incremental perception of the GA as they emerge overtime. In this paper, we propose an ontology based GA recognition where three inference Hidden Markov Models (HMM's) are used for predicting group activities taking place in outdoor environments and different task operational taxonomy. The three competing models include: a concatenated HMM, a cascaded HMM, and a context-based HMM. The proposed ontology based GA-HMM was subjected to set of semantically annotated visual observations from outdoor group activity experiments. Experimental results from GA-HMM are presented with technical discussions on design of each model and their potential implication to Persistent Surveillance Systems (PSS).

AB - Understanding of Group Activities (GA) has significant applications in civilian and military domains. The process of understanding GA is typically involved with spatiotemporal analysis of multi-modality sensor data. Video imagery is one popular sensing modality that offers rich data, however, data associated with imagery source may become fragmented and discontinued due to a number of reasons (e.g., data transmission, or observation obstructions and occlusions). However, making sense out of video imagery is a real challenge. It requires a proper inference working model capable of analyzing video imagery frame by frame, extract and inference spatiotemporal information pertaining to observations while developing an incremental perception of the GA as they emerge overtime. In this paper, we propose an ontology based GA recognition where three inference Hidden Markov Models (HMM's) are used for predicting group activities taking place in outdoor environments and different task operational taxonomy. The three competing models include: a concatenated HMM, a cascaded HMM, and a context-based HMM. The proposed ontology based GA-HMM was subjected to set of semantically annotated visual observations from outdoor group activity experiments. Experimental results from GA-HMM are presented with technical discussions on design of each model and their potential implication to Persistent Surveillance Systems (PSS).

UR - http://www.scopus.com/inward/record.url?scp=84906243338&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84906243338&partnerID=8YFLogxK

U2 - 10.1117/12.2050909

DO - 10.1117/12.2050909

M3 - Conference contribution

SN - 9781628410280

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII

PB - SPIE

ER -

Elangovan V, Shirkhodaie A. Knowledge discovery in group activities through sequential observation analysis. In Signal Processing, Sensor/Information Fusion, and Target Recognition XXIII. SPIE. 2014. 90910T. (Proceedings of SPIE - The International Society for Optical Engineering). https://doi.org/10.1117/12.2050909