Personalised modelling of facial action unit intensity

Shuang Yang; Ognjen Rudovic; Vladimir Pavlovic; Maja  Pantic

doi:10.1007/978-3-319-14364-4_26

Personalised modelling of facial action unit intensity

Shuang Yang
, Ognjen Rudovic
, Vladimir Pavlovic
, Maja Pantic

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

16 Citations (Scopus)

11 Downloads (Pure)

Abstract

Facial expressions depend greatly on facial morphology and expressiveness of the observed person. Recent studies have shown great improvement of the personalized over non-personalized models in variety of facial expression related tasks, such as face and emotion recognition. However, in the context of facial action unit (AU) intensity estimation, personalized modeling has been scarcely investigated. In this paper, we propose a two-step approach for personalized modeling of facial AU intensity from spontaneously displayed facial expressions. In the first step, we perform facial feature decomposition using the proposed matrix decomposition algorithm that separates the person’s identity from facial expression. These two are then jointly modeled using the framework of Conditional Ordinal Random Fields, resulting in a personalized model for intensity estimation of AUs. Our experimental results show that the proposed personalized model largely outperforms non-personalized models for intensity estimation of AUs.

Original language	English
Title of host publication	Advances in Visual Computing
Subtitle of host publication	10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings
Editors	George Bebis, Richard Boyle, Bahram Parvin, Darko Koracin
Place of Publication	Cham
Publisher	Springer
Pages	269-281
Number of pages	13
Volume	Part II
ISBN (Electronic)	978-3-319-14364-4
ISBN (Print)	978-3-319-14363-7
DOIs	https://doi.org/10.1007/978-3-319-14364-4_26
Publication status	Published - Dec 2014
Event	10th International Symposium on Visual Computing, ISVC 2014 - Las Vegas, United States Duration: 8 Dec 2014 → 10 Dec 2014 Conference number: 10

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	8888
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	10th International Symposium on Visual Computing, ISVC 2014
Abbreviated title	ISVC
Country/Territory	United States
City	Las Vegas
Period	8/12/14 → 10/12/14

Keywords

HMI-HF: Human Factors
EWI-25809
EC Grant Agreement nr.: FP7/611153
EC Grant Agreement nr.: FP7/2007-2013
METIS-309928
IR-94676

Access to Document

10.1007/978-3-319-14364-4_26

Cite this

Yang, S., Rudovic, O., Pavlovic, V., & Pantic, M. (2014). Personalised modelling of facial action unit intensity. In G. Bebis, R. Boyle, B. Parvin, & D. Koracin (Eds.), Advances in Visual Computing: 10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings (Vol. Part II, pp. 269-281). (Lecture Notes in Computer Science; Vol. 8888). Springer. https://doi.org/10.1007/978-3-319-14364-4_26

Yang, Shuang ; Rudovic, Ognjen ; Pavlovic, Vladimir et al. / Personalised modelling of facial action unit intensity. Advances in Visual Computing: 10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings. editor / George Bebis ; Richard Boyle ; Bahram Parvin ; Darko Koracin. Vol. Part II Cham : Springer, 2014. pp. 269-281 (Lecture Notes in Computer Science).

@inproceedings{b556d617c2c74c66bb04878c07ece532,

title = "Personalised modelling of facial action unit intensity",

abstract = "Facial expressions depend greatly on facial morphology and expressiveness of the observed person. Recent studies have shown great improvement of the personalized over non-personalized models in variety of facial expression related tasks, such as face and emotion recognition. However, in the context of facial action unit (AU) intensity estimation, personalized modeling has been scarcely investigated. In this paper, we propose a two-step approach for personalized modeling of facial AU intensity from spontaneously displayed facial expressions. In the first step, we perform facial feature decomposition using the proposed matrix decomposition algorithm that separates the person{\textquoteright}s identity from facial expression. These two are then jointly modeled using the framework of Conditional Ordinal Random Fields, resulting in a personalized model for intensity estimation of AUs. Our experimental results show that the proposed personalized model largely outperforms non-personalized models for intensity estimation of AUs.",

keywords = "HMI-HF: Human Factors, EWI-25809, EC Grant Agreement nr.: FP7/611153, EC Grant Agreement nr.: FP7/2007-2013, METIS-309928, IR-94676",

author = "Shuang Yang and Ognjen Rudovic and Vladimir Pavlovic and Maja Pantic",

note = "eemcs-eprint-25809 ; 10th International Symposium on Visual Computing, ISVC 2014, ISVC ; Conference date: 08-12-2014 Through 10-12-2014",

year = "2014",

month = dec,

doi = "10.1007/978-3-319-14364-4\_26",

language = "English",

isbn = "978-3-319-14363-7",

volume = "Part II",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "269--281",

editor = "George Bebis and Richard Boyle and Bahram Parvin and Darko Koracin",

booktitle = "Advances in Visual Computing",

address = "Germany",

}

Yang, S, Rudovic, O, Pavlovic, V & Pantic, M 2014, Personalised modelling of facial action unit intensity. in G Bebis, R Boyle, B Parvin & D Koracin (eds), Advances in Visual Computing: 10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings. vol. Part II, Lecture Notes in Computer Science, vol. 8888, Springer, Cham, pp. 269-281, 10th International Symposium on Visual Computing, ISVC 2014, Las Vegas, Nevada, United States, 8/12/14. https://doi.org/10.1007/978-3-319-14364-4_26

Personalised modelling of facial action unit intensity. / Yang, Shuang; Rudovic, Ognjen; Pavlovic, Vladimir et al.
Advances in Visual Computing: 10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings. ed. / George Bebis; Richard Boyle; Bahram Parvin; Darko Koracin. Vol. Part II Cham: Springer, 2014. p. 269-281 (Lecture Notes in Computer Science; Vol. 8888).

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

TY - GEN

T1 - Personalised modelling of facial action unit intensity

AU - Yang, Shuang

AU - Rudovic, Ognjen

AU - Pavlovic, Vladimir

AU - Pantic, Maja

N1 - Conference code: 10

PY - 2014/12

Y1 - 2014/12

N2 - Facial expressions depend greatly on facial morphology and expressiveness of the observed person. Recent studies have shown great improvement of the personalized over non-personalized models in variety of facial expression related tasks, such as face and emotion recognition. However, in the context of facial action unit (AU) intensity estimation, personalized modeling has been scarcely investigated. In this paper, we propose a two-step approach for personalized modeling of facial AU intensity from spontaneously displayed facial expressions. In the first step, we perform facial feature decomposition using the proposed matrix decomposition algorithm that separates the person’s identity from facial expression. These two are then jointly modeled using the framework of Conditional Ordinal Random Fields, resulting in a personalized model for intensity estimation of AUs. Our experimental results show that the proposed personalized model largely outperforms non-personalized models for intensity estimation of AUs.

AB - Facial expressions depend greatly on facial morphology and expressiveness of the observed person. Recent studies have shown great improvement of the personalized over non-personalized models in variety of facial expression related tasks, such as face and emotion recognition. However, in the context of facial action unit (AU) intensity estimation, personalized modeling has been scarcely investigated. In this paper, we propose a two-step approach for personalized modeling of facial AU intensity from spontaneously displayed facial expressions. In the first step, we perform facial feature decomposition using the proposed matrix decomposition algorithm that separates the person’s identity from facial expression. These two are then jointly modeled using the framework of Conditional Ordinal Random Fields, resulting in a personalized model for intensity estimation of AUs. Our experimental results show that the proposed personalized model largely outperforms non-personalized models for intensity estimation of AUs.

KW - HMI-HF: Human Factors

KW - EWI-25809

KW - EC Grant Agreement nr.: FP7/611153

KW - EC Grant Agreement nr.: FP7/2007-2013

KW - METIS-309928

KW - IR-94676

U2 - 10.1007/978-3-319-14364-4_26

DO - 10.1007/978-3-319-14364-4_26

M3 - Conference contribution

SN - 978-3-319-14363-7

VL - Part II

T3 - Lecture Notes in Computer Science

SP - 269

EP - 281

BT - Advances in Visual Computing

A2 - Bebis, George

A2 - Boyle, Richard

A2 - Parvin, Bahram

A2 - Koracin, Darko

PB - Springer

CY - Cham

T2 - 10th International Symposium on Visual Computing, ISVC 2014

Y2 - 8 December 2014 through 10 December 2014

ER -

Yang S, Rudovic O, Pavlovic V, Pantic M. Personalised modelling of facial action unit intensity. In Bebis G, Boyle R, Parvin B, Koracin D, editors, Advances in Visual Computing: 10th International Symposium, ISVC 2014, Las Vegas, NV, USA, December 8-10, 2014, Proceedings. Vol. Part II. Cham: Springer. 2014. p. 269-281. (Lecture Notes in Computer Science). doi: 10.1007/978-3-319-14364-4_26

Personalised modelling of facial action unit intensity

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this