Spatiotemporal Facial Super-Pixels for Pain Detection

Dennis Holm Lundtoft; Kamal Nasrollahi; Thomas B. Moeslund; Sergio Escalera Guerrero

doi:10.1007/978-3-319-41778-3_4

Spatiotemporal Facial Super-Pixels for Pain Detection

Dennis Holm Lundtoft, Kamal Nasrollahi, Thomas B. Moeslund, Sergio Escalera Guerrero

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

453 Downloads (Pure)

Abstract

Pain detection using facial images is of critical importance in many eHealth applications. Since pain is a spatiotemporal process, recent works on this topic employ facial spatiotemporal features to detect pain. These systems extract such features from the entire area of the face. In this paper, we show that by employing super-pixels we can divide the face into three regions, in a way that only one of these regions (about one third of the face) contributes to the pain estimation and the other two regions can be discarded. The experimental results on the UNBC- McMaster database show that the proposed system using this single region outperforms state-of-the-art systems in detecting no-pain scenarios, while it reaches comparable results in detecting weak and severe pain scenarios.

Original language	English
Title of host publication	IX Conference on Articulated Motion and Deformable Objects
Editors	Francisco José Perales, Josef Kittler
Place of Publication	Spain
Publisher	Springer
Publication date	2 Jul 2016
Pages	34-43
DOIs	https://doi.org/10.1007/978-3-319-41778-3_4
Publication status	Published - 2 Jul 2016
Event	IX Conference on Articulated Motion and Deformable Objects - Mallorca, Spain Duration: 13 Jul 2016 → 15 Jul 2016 http://amdo2016.uib.es/

Conference

Conference	IX Conference on Articulated Motion and Deformable Objects
Country/Territory	Spain
City	Mallorca
Period	13/07/2016 → 15/07/2016
Internet address	http://amdo2016.uib.es/

Series	Lecture Notes in Computer Science
Volume	9756
ISSN	0302-9743

Keywords

facial images
super-pixels
spatiotemporal filters
pain detection

Access to Document

10.1007/978-3-319-41778-3_4

pre-print

AUB Link

Search for the material in Aalborg University Library's search engine

Cite this

@inproceedings{b7dae0ebb19742e2ad735652ba6d2e86,

title = "Spatiotemporal Facial Super-Pixels for Pain Detection",

abstract = "Pain detection using facial images is of critical importance in many eHealth applications. Since pain is a spatiotemporal process, recent works on this topic employ facial spatiotemporal features to detect pain. These systems extract such features from the entire area of the face. In this paper, we show that by employing super-pixels we can divide the face into three regions, in a way that only one of these regions (about one third of the face) contributes to the pain estimation and the other two regions can be discarded. The experimental results on the UNBC- McMaster database show that the proposed system using this single region outperforms state-of-the-art systems in detecting no-pain scenarios, while it reaches comparable results in detecting weak and severe pain scenarios.",

keywords = "facial images, super-pixels, spatiotemporal filters, pain detection",

author = "Lundtoft, {Dennis Holm} and Kamal Nasrollahi and Moeslund, {Thomas B.} and Guerrero, {Sergio Escalera}",

year = "2016",

month = jul,

day = "2",

doi = "10.1007/978-3-319-41778-3_4",

language = "English",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "34--43",

editor = "Perales, {Francisco Jos{\'e}} and Josef Kittler",

booktitle = "IX Conference on Articulated Motion and Deformable Objects",

address = "Germany",

note = "IX Conference on Articulated Motion and Deformable Objects, AMDO 2016 ; Conference date: 13-07-2016 Through 15-07-2016",

url = "http://amdo2016.uib.es/",

}

Lundtoft, DH, Nasrollahi, K , Moeslund, TB & Guerrero, SE 2016, Spatiotemporal Facial Super-Pixels for Pain Detection. in FJ Perales & J Kittler (eds), IX Conference on Articulated Motion and Deformable Objects. Springer, Spain, Lecture Notes in Computer Science, vol. 9756, pp. 34-43, IX Conference on Articulated Motion and Deformable Objects, Mallorca, Spain, 13/07/2016. https://doi.org/10.1007/978-3-319-41778-3_4

Spatiotemporal Facial Super-Pixels for Pain Detection. / Lundtoft, Dennis Holm; Nasrollahi, Kamal ; Moeslund, Thomas B. et al.
IX Conference on Articulated Motion and Deformable Objects. ed. / Francisco José Perales; Josef Kittler. Spain: Springer, 2016. p. 34-43 (Lecture Notes in Computer Science, Vol. 9756).

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

TY - GEN

T1 - Spatiotemporal Facial Super-Pixels for Pain Detection

AU - Lundtoft, Dennis Holm

AU - Nasrollahi, Kamal

AU - Moeslund, Thomas B.

AU - Guerrero, Sergio Escalera

PY - 2016/7/2

Y1 - 2016/7/2

N2 - Pain detection using facial images is of critical importance in many eHealth applications. Since pain is a spatiotemporal process, recent works on this topic employ facial spatiotemporal features to detect pain. These systems extract such features from the entire area of the face. In this paper, we show that by employing super-pixels we can divide the face into three regions, in a way that only one of these regions (about one third of the face) contributes to the pain estimation and the other two regions can be discarded. The experimental results on the UNBC- McMaster database show that the proposed system using this single region outperforms state-of-the-art systems in detecting no-pain scenarios, while it reaches comparable results in detecting weak and severe pain scenarios.

AB - Pain detection using facial images is of critical importance in many eHealth applications. Since pain is a spatiotemporal process, recent works on this topic employ facial spatiotemporal features to detect pain. These systems extract such features from the entire area of the face. In this paper, we show that by employing super-pixels we can divide the face into three regions, in a way that only one of these regions (about one third of the face) contributes to the pain estimation and the other two regions can be discarded. The experimental results on the UNBC- McMaster database show that the proposed system using this single region outperforms state-of-the-art systems in detecting no-pain scenarios, while it reaches comparable results in detecting weak and severe pain scenarios.

KW - facial images

KW - super-pixels

KW - spatiotemporal filters

KW - pain detection

U2 - 10.1007/978-3-319-41778-3_4

DO - 10.1007/978-3-319-41778-3_4

M3 - Article in proceeding

T3 - Lecture Notes in Computer Science

SP - 34

EP - 43

BT - IX Conference on Articulated Motion and Deformable Objects

A2 - Perales, Francisco José

A2 - Kittler, Josef

PB - Springer

CY - Spain

T2 - IX Conference on Articulated Motion and Deformable Objects

Y2 - 13 July 2016 through 15 July 2016

ER -

Spatiotemporal Facial Super-Pixels for Pain Detection

Abstract

Conference

Keywords

Access to Document

AUB Link

Fingerprint

Cite this