A Hybrid System for On-line Blink Detection

Yijia Sun; Stefanos Zafeiriou; Maja  Pantic

A Hybrid System for On-line Blink Detection

Yijia Sun, Stefanos Zafeiriou, Maja Pantic

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

93 Downloads (Pure)

Abstract

Eye blinking behaviour has been shown to be one of the most informative non-verbal behavioural cues for indicating deceptive behaviour. Traditional blink detection methods tend to use a tracker to extract static eye region images and classify those images as open and closed eyes in order to detect blinks. However, those recognition systems are frame based and do not incorporate temporal information. For this reason, they perform poorly as the tracker fails to detect eyes due to rapid head movement or occlusion. In this paper, we present an approach which combines Hidden Markov Models and Support Vector Machines to model the temporal dynamics of eye blinks and improve the blink detection accuracy.

Original language	English
Title of host publication	Proceedings of the Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium
Subtitle of host publication	January 2013
Editors	Matthew Jensen, Thomas Meservy, Judee Burgoon, Jay Nunamaker
Number of pages	5
Publication status	Published - 7 Jan 2013
Event	HICSS-46 Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium 2013 - Maui, United States Duration: 7 Jan 2013 → 8 Jan 2013

Conference

Conference	HICSS-46 Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium 2013
Country/Territory	United States
City	Maui
Period	7/01/13 → 8/01/13

Access to Document

Sun2013hybridFinal published version, 144 KB

Cite this

@inproceedings{b781626a4cbc47b1b43cd712e2309e1f,

title = "A Hybrid System for On-line Blink Detection",

abstract = "Eye blinking behaviour has been shown to be one of the most informative non-verbal behavioural cues for indicating deceptive behaviour. Traditional blink detection methods tend to use a tracker to extract static eye region images and classify those images as open and closed eyes in order to detect blinks. However, those recognition systems are frame based and do not incorporate temporal information. For this reason, they perform poorly as the tracker fails to detect eyes due to rapid head movement or occlusion. In this paper, we present an approach which combines Hidden Markov Models and Support Vector Machines to model the temporal dynamics of eye blinks and improve the blink detection accuracy.",

author = "Yijia Sun and Stefanos Zafeiriou and Maja Pantic",

year = "2013",

month = jan,

day = "7",

language = "English",

editor = "Matthew Jensen and Thomas Meservy and Judee Burgoon and Jay Nunamaker",

booktitle = "Proceedings of the Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium",

note = "HICSS-46 Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium 2013 ; Conference date: 07-01-2013 Through 08-01-2013",

}

Sun, Y, Zafeiriou, S & Pantic, M 2013, A Hybrid System for On-line Blink Detection. in M Jensen, T Meservy, J Burgoon & J Nunamaker (eds), Proceedings of the Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium: January 2013. HICSS-46 Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium 2013, Maui, Hawaii, United States, 7/01/13.

A Hybrid System for On-line Blink Detection. / Sun, Yijia; Zafeiriou, Stefanos; Pantic, Maja .
Proceedings of the Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium: January 2013. ed. / Matthew Jensen; Thomas Meservy; Judee Burgoon; Jay Nunamaker. 2013.

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

TY - GEN

T1 - A Hybrid System for On-line Blink Detection

AU - Sun, Yijia

AU - Zafeiriou, Stefanos

AU - Pantic, Maja

PY - 2013/1/7

Y1 - 2013/1/7

N2 - Eye blinking behaviour has been shown to be one of the most informative non-verbal behavioural cues for indicating deceptive behaviour. Traditional blink detection methods tend to use a tracker to extract static eye region images and classify those images as open and closed eyes in order to detect blinks. However, those recognition systems are frame based and do not incorporate temporal information. For this reason, they perform poorly as the tracker fails to detect eyes due to rapid head movement or occlusion. In this paper, we present an approach which combines Hidden Markov Models and Support Vector Machines to model the temporal dynamics of eye blinks and improve the blink detection accuracy.

AB - Eye blinking behaviour has been shown to be one of the most informative non-verbal behavioural cues for indicating deceptive behaviour. Traditional blink detection methods tend to use a tracker to extract static eye region images and classify those images as open and closed eyes in order to detect blinks. However, those recognition systems are frame based and do not incorporate temporal information. For this reason, they perform poorly as the tracker fails to detect eyes due to rapid head movement or occlusion. In this paper, we present an approach which combines Hidden Markov Models and Support Vector Machines to model the temporal dynamics of eye blinks and improve the blink detection accuracy.

M3 - Conference contribution

BT - Proceedings of the Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium

A2 - Jensen, Matthew

A2 - Meservy, Thomas

A2 - Burgoon, Judee

A2 - Nunamaker, Jay

T2 - HICSS-46 Rapid Screening Technologies, Deception Detection and Credibility Assessment Symposium 2013

Y2 - 7 January 2013 through 8 January 2013

ER -

A Hybrid System for On-line Blink Detection

Abstract

Conference

Access to Document

Fingerprint

Cite this