Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions

Damien Dupré; Nicole Andelic; Gawain Morrison; Gary McKeown

doi:10.1109/PERCOMW.2018.8480127

Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions

Damien Dupré, Nicole Andelic, Gawain Morrison, Gary McKeown

Economics

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

15 Citations (Scopus)

Abstract

Automatic facial expression recognition systems can provide information about our emotions and how they change over time. However, based on different statistical methods the results of automatic systems have not yet been compared. In the current paper we evaluate the emotion detection between three different commercial systems (i.e.Affectiva, Kairos and Microsoft) when detecting dynamic and spontaneous facial expressions. Even if the study was performed on a limited sample of videos, the results show significant differences between the systems for the same video and per system across comparable facial expressions. Finally, we reflect on the implications according the generalization of the results provided by automatic emotion detection.

Original language	English
Title of host publication	Pervasive Computing and Communications Workshops (PerCom)
Subtitle of host publication	IEEE Annual Conference
Publisher	IEEE Press
Number of pages	6
ISBN (Print)	978-1-5386-3228-4
DOIs	https://doi.org/10.1109/PERCOMW.2018.8480127
Publication status	Published - 8 Oct 2018
Event	2018 IEEE International Conference on Pervasive Computing and Communications Workshops : PerCom Workshops 2018 - Athens, Greece Duration: 19 Mar 2018 → 23 Mar 2018

Conference

Conference	2018 IEEE International Conference on Pervasive Computing and Communications Workshops
Country/Territory	Greece
City	Athens
Period	19/03/18 → 23/03/18

Keywords

emotion
facial expression
automatic recognition

Access to Document

10.1109/PERCOMW.2018.8480127Licence: Unspecified

Cite this

Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions. / Dupré, Damien; Andelic, Nicole; Morrison, Gawain et al.

Pervasive Computing and Communications Workshops (PerCom): IEEE Annual Conference. IEEE Press, 2018.

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

Dupré, D, Andelic, N, Morrison, G & McKeown, G 2018, Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions. in Pervasive Computing and Communications Workshops (PerCom): IEEE Annual Conference. IEEE Press, 2018 IEEE International Conference on Pervasive Computing and Communications Workshops , Athens, Greece, 19/03/18. https://doi.org/10.1109/PERCOMW.2018.8480127

@inbook{f49a9f48d4494b6ca9d75480cbb1bfd4,

title = "Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions",

abstract = "Automatic facial expression recognition systems can provide information about our emotions and how they change over time. However, based on different statistical methods the results of automatic systems have not yet been compared. In the current paper we evaluate the emotion detection between three different commercial systems (i.e.Affectiva, Kairos and Microsoft) when detecting dynamic and spontaneous facial expressions. Even if the study was performed on a limited sample of videos, the results show significant differences between the systems for the same video and per system across comparable facial expressions. Finally, we reflect on the implications according the generalization of the results provided by automatic emotion detection.",

keywords = "emotion, facial expression, automatic recognition",

author = "Damien Dupr{\'e} and Nicole Andelic and Gawain Morrison and Gary McKeown",

note = "We thank the representatives of Affectiva, Kairos and Microsoft who agreed for the communication of the data presented in this paper. ; 2018 IEEE International Conference on Pervasive Computing and Communications Workshops : PerCom Workshops 2018 ; Conference date: 19-03-2018 Through 23-03-2018",

year = "2018",

month = oct,

day = "8",

doi = "10.1109/PERCOMW.2018.8480127",

language = "English",

isbn = "978-1-5386-3228-4",

booktitle = "Pervasive Computing and Communications Workshops (PerCom)",

publisher = "IEEE Press",

}

TY - CHAP

T1 - Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions

AU - Dupré, Damien

AU - Andelic, Nicole

AU - Morrison, Gawain

AU - McKeown, Gary

N1 - We thank the representatives of Affectiva, Kairos and Microsoft who agreed for the communication of the data presented in this paper.

PY - 2018/10/8

Y1 - 2018/10/8

N2 - Automatic facial expression recognition systems can provide information about our emotions and how they change over time. However, based on different statistical methods the results of automatic systems have not yet been compared. In the current paper we evaluate the emotion detection between three different commercial systems (i.e.Affectiva, Kairos and Microsoft) when detecting dynamic and spontaneous facial expressions. Even if the study was performed on a limited sample of videos, the results show significant differences between the systems for the same video and per system across comparable facial expressions. Finally, we reflect on the implications according the generalization of the results provided by automatic emotion detection.

AB - Automatic facial expression recognition systems can provide information about our emotions and how they change over time. However, based on different statistical methods the results of automatic systems have not yet been compared. In the current paper we evaluate the emotion detection between three different commercial systems (i.e.Affectiva, Kairos and Microsoft) when detecting dynamic and spontaneous facial expressions. Even if the study was performed on a limited sample of videos, the results show significant differences between the systems for the same video and per system across comparable facial expressions. Finally, we reflect on the implications according the generalization of the results provided by automatic emotion detection.

KW - emotion

KW - facial expression

KW - automatic recognition

U2 - 10.1109/PERCOMW.2018.8480127

DO - 10.1109/PERCOMW.2018.8480127

M3 - Chapter (peer-reviewed)

SN - 978-1-5386-3228-4

BT - Pervasive Computing and Communications Workshops (PerCom)

PB - IEEE Press

T2 - 2018 IEEE International Conference on Pervasive Computing and Communications Workshops

Y2 - 19 March 2018 through 23 March 2018

ER -

Accuracy of three commercial automatic emotion recognition systems across different individuals and their facial expressions

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this