P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions

Facial expressions are one of the most important means of nonverbal communication transporting both emotional and conversational content. For investigating this large space of expressions we recently developed a large database containing dynamic emotional and conversational expressions in Germany (M...

Full description

Bibliographic Details
Main Authors: Haenah Lee, Ahyoung Shin, BoRa Kim, Christian Wallraven
Format: Article
Language:English
Published: SAGE Publishing 2012-10-01
Series:i-Perception
Online Access:https://doi.org/10.1068/if694
id doaj-57825fc6be744234bed4774735689a8b
record_format Article
spelling doaj-57825fc6be744234bed4774735689a8b2020-11-25T03:03:15ZengSAGE Publishingi-Perception2041-66952012-10-01310.1068/if69410.1068_if694P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational ExpressionsHaenah Lee0Ahyoung ShinBoRa KimChristian WallravenKorea University, KoreaFacial expressions are one of the most important means of nonverbal communication transporting both emotional and conversational content. For investigating this large space of expressions we recently developed a large database containing dynamic emotional and conversational expressions in Germany (MPI facial expression database). As facial expressions crucially depend on the cultural context, however, a similar resource is needed for studies outside of Germany. Here, we introduce and validate a new, extensive Korean facial expression database containing dynamic emotional and conversational information. Ten individuals performed 62 expressions following a method-acting protocol, in which each person was asked to imagine themselves in one of 62 corresponding everyday scenarios and to react accordingly. To validate this database, we conducted two experiments: 20 participants were asked to name the appropriate expression for each of the 62 everyday scenarios shown as text. Ten additional participants were asked to name each of the 62 expression videos from 10 actors in addition to rating its naturalness. All naming answers were then rated as valid or invalid. Scenario validation yielded 89% valid answers showing that the scenarios are effective in eliciting appropriate expressions. Video sequences were judged as natural with an average of 66% valid answers. This is an excellent result considering that videos were seen without any conversational context and that 62 expressions were to be recognized. These results validate our Korean database and, as they also parallel the German validation results, will enable detailed cross-cultural comparisons of the complex space of emotional and conversational expressions.https://doi.org/10.1068/if694
collection DOAJ
language English
format Article
sources DOAJ
author Haenah Lee
Ahyoung Shin
BoRa Kim
Christian Wallraven
spellingShingle Haenah Lee
Ahyoung Shin
BoRa Kim
Christian Wallraven
P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
i-Perception
author_facet Haenah Lee
Ahyoung Shin
BoRa Kim
Christian Wallraven
author_sort Haenah Lee
title P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
title_short P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
title_full P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
title_fullStr P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
title_full_unstemmed P2-35: The KU Facial Expression Database: A Validated Database of Emotional and Conversational Expressions
title_sort p2-35: the ku facial expression database: a validated database of emotional and conversational expressions
publisher SAGE Publishing
series i-Perception
issn 2041-6695
publishDate 2012-10-01
description Facial expressions are one of the most important means of nonverbal communication transporting both emotional and conversational content. For investigating this large space of expressions we recently developed a large database containing dynamic emotional and conversational expressions in Germany (MPI facial expression database). As facial expressions crucially depend on the cultural context, however, a similar resource is needed for studies outside of Germany. Here, we introduce and validate a new, extensive Korean facial expression database containing dynamic emotional and conversational information. Ten individuals performed 62 expressions following a method-acting protocol, in which each person was asked to imagine themselves in one of 62 corresponding everyday scenarios and to react accordingly. To validate this database, we conducted two experiments: 20 participants were asked to name the appropriate expression for each of the 62 everyday scenarios shown as text. Ten additional participants were asked to name each of the 62 expression videos from 10 actors in addition to rating its naturalness. All naming answers were then rated as valid or invalid. Scenario validation yielded 89% valid answers showing that the scenarios are effective in eliciting appropriate expressions. Video sequences were judged as natural with an average of 66% valid answers. This is an excellent result considering that videos were seen without any conversational context and that 62 expressions were to be recognized. These results validate our Korean database and, as they also parallel the German validation results, will enable detailed cross-cultural comparisons of the complex space of emotional and conversational expressions.
url https://doi.org/10.1068/if694
work_keys_str_mv AT haenahlee p235thekufacialexpressiondatabaseavalidateddatabaseofemotionalandconversationalexpressions
AT ahyoungshin p235thekufacialexpressiondatabaseavalidateddatabaseofemotionalandconversationalexpressions
AT borakim p235thekufacialexpressiondatabaseavalidateddatabaseofemotionalandconversationalexpressions
AT christianwallraven p235thekufacialexpressiondatabaseavalidateddatabaseofemotionalandconversationalexpressions
_version_ 1724686788304306176