Signals of Emotion Regulation in a Social Dilemma: Detection from Face and Context (bibtex)
by Hoegen, Rens, Gratch, Jonathan, Parkinson, Brian and Shore, Danielle
Abstract:
In social decision-making tasks, facial expressions are informative signals that indicate motives and intentions. As people are aware that their expressions influence partner behavior, expressions may be strategically regulated in competitive environments to influence a social partner’s decisionmaking. In this work, we examine facial expressions and their strategic regulation within the context of an iterated prisoner’s dilemma. Utilizing video-cued rating procedures, we examine several key questions about the functionality of facial expressions in social decision-making. First, we assess the extent to which emotion and expression regulation are accurately detected from dynamic facial expressions in interpersonal interactions. Second, we explore which facial cues are utilized to evaluate emotion and regulation information. Finally, we investigate the role of context in participants’ emotion and regulation judgments. Results show that participants accurately perceive facial emotion and expression regulation, although they are better at recognizing emotions than regulation. Using automated expression analysis and stepwise regression, we constructed models that use action units from participant videos to predict their video-cued emotion and regulation ratings. We show that these models perform similarly and, in some cases, better than participants do. Moreover, these models demonstrate that game state information improves predictive accuracy, thus implying that context information is important in the evaluation of facial expressions.
Reference:
Signals of Emotion Regulation in a Social Dilemma: Detection from Face and Context (Hoegen, Rens, Gratch, Jonathan, Parkinson, Brian and Shore, Danielle), In Proceedings of the 8th International Conference on Affective Computing and Intelligent Interaction (ACII) roceedings of the, IEEE, 2019.
Bibtex Entry:
@inproceedings{hoegen_signals_2019,
	address = {Cambridge, UK},
	title = {Signals of {Emotion} {Regulation} in a {Social} {Dilemma}: {Detection} from {Face} and {Context}},
	url = {https://ieeexplore.ieee.org/xpl/conhome/8911251/proceeding},
	abstract = {In social decision-making tasks, facial expressions are informative signals that indicate motives and intentions. As people are aware that their expressions influence partner behavior, expressions may be strategically regulated in competitive environments to influence a social partner’s decisionmaking. In this work, we examine facial expressions and their strategic regulation within the context of an iterated prisoner’s dilemma. Utilizing video-cued rating procedures, we examine several key questions about the functionality of facial expressions in social decision-making. First, we assess the extent to which emotion and expression regulation are accurately detected from dynamic facial expressions in interpersonal interactions. Second, we explore which facial cues are utilized to evaluate emotion and regulation information. Finally, we investigate the role of context in participants’ emotion and regulation judgments. Results show that participants accurately perceive facial emotion and expression regulation, although they are better at recognizing emotions than regulation. Using automated expression analysis and stepwise regression, we constructed models that use action units from participant videos to predict their video-cued emotion and regulation ratings. We show that these models perform similarly and, in some cases, better than participants do. Moreover, these models demonstrate that game state information improves predictive accuracy, thus implying that context information is important in the evaluation of facial expressions.},
	booktitle = {Proceedings of the 8th {International} {Conference} on {Affective} {Computing} and {Intelligent} {Interaction} ({ACII}) roceedings of the},
	publisher = {IEEE},
	author = {Hoegen, Rens and Gratch, Jonathan and Parkinson, Brian and Shore, Danielle},
	month = sep,
	year = {2019},
	keywords = {Virtual Humans},
	pages = {7}
}
Powered by bibtexbrowser