3D face tracking for infant monitoring using dense HOG and drift reduction

R.W.J.J. Saeijs; W.E. Tjon A Ten; P.H.N. de With

3D face tracking for infant monitoring using dense HOG and drift reduction

R.W.J.J. Saeijs
, W.E. Tjon A Ten
, P.H.N. de With

Architectures for Reliable Image Analysis

Research output: Contribution to conference › Paper › Academic

109 Downloads (Pure)

Abstract

This paper presents a new algorithm for 3D face tracking intended for clinical infant pain monitoring. The algorithm uses a cylinder head model and 3D head pose recovery by alignment of dynamically extracted templates based on dense-HOG features. Drift reduction is obtained from re-registration in combination with multi-pose state estimation by a square-root unscented Kalman filter. Results on videos of moving infants in hospital show good tracking for poses up to 50 degrees from upright-frontal, with mean eyelocation error relative to inter-ocular distance below 9%.

Original language	English
Pages	1-2
Publication status	Published - 2016
Event	The Netherlands Conference on Computer Vision (NCCV 2016) - Lunteren, Netherlands Duration: 12 Dec 2016 → 13 Dec 2016 http://www.nccv16.nl

Conference

Conference	The Netherlands Conference on Computer Vision (NCCV 2016)
Abbreviated title	NCCV 2016
Country/Territory	Netherlands
City	Lunteren
Period	12/12/16 → 13/12/16
Internet address	http://www.nccv16.nl

Keywords

face tracking
pain monitoring
cylinder head model
dense HOG

Access to Document

NCCV2016_3d-face-tracking-etc_Saeijs_TjonATen_deWith_FINALFinal author version , 779 KBLicence: CC BY-SA

Cite this

@conference{c98706bc9e6f4a3887ba9c0da6adb8f0,

title = "3D face tracking for infant monitoring using dense HOG and drift reduction",

abstract = "This paper presents a new algorithm for 3D face tracking intended for clinical infant pain monitoring. The algorithm uses a cylinder head model and 3D head pose recovery by alignment of dynamically extracted templates based on dense-HOG features. Drift reduction is obtained from re-registration in combination with multi-pose state estimation by a square-root unscented Kalman filter. Results on videos of moving infants in hospital show good tracking for poses up to 50 degrees from upright-frontal, with mean eyelocation error relative to inter-ocular distance below 9\%.",

keywords = "face tracking, pain monitoring, cylinder head model, dense HOG",

author = "R.W.J.J. Saeijs and \{Tjon A Ten\}, W.E. and \{de With\}, P.H.N.",

year = "2016",

language = "English",

pages = "1--2",

note = "The Netherlands Conference on Computer Vision (NCCV 2016), NCCV 2016 ; Conference date: 12-12-2016 Through 13-12-2016",

url = "http://www.nccv16.nl",

}

TY - CONF

T1 - 3D face tracking for infant monitoring using dense HOG and drift reduction

AU - Saeijs, R.W.J.J.

AU - Tjon A Ten, W.E.

AU - de With, P.H.N.

PY - 2016

Y1 - 2016

N2 - This paper presents a new algorithm for 3D face tracking intended for clinical infant pain monitoring. The algorithm uses a cylinder head model and 3D head pose recovery by alignment of dynamically extracted templates based on dense-HOG features. Drift reduction is obtained from re-registration in combination with multi-pose state estimation by a square-root unscented Kalman filter. Results on videos of moving infants in hospital show good tracking for poses up to 50 degrees from upright-frontal, with mean eyelocation error relative to inter-ocular distance below 9%.

AB - This paper presents a new algorithm for 3D face tracking intended for clinical infant pain monitoring. The algorithm uses a cylinder head model and 3D head pose recovery by alignment of dynamically extracted templates based on dense-HOG features. Drift reduction is obtained from re-registration in combination with multi-pose state estimation by a square-root unscented Kalman filter. Results on videos of moving infants in hospital show good tracking for poses up to 50 degrees from upright-frontal, with mean eyelocation error relative to inter-ocular distance below 9%.

KW - face tracking

KW - pain monitoring

KW - cylinder head model

KW - dense HOG

M3 - Paper

SP - 1

EP - 2

T2 - The Netherlands Conference on Computer Vision (NCCV 2016)

Y2 - 12 December 2016 through 13 December 2016

ER -

3D face tracking for infant monitoring using dense HOG and drift reduction

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this