Introducing the Geneva Multimodal expression corpus for experimental research on emotion perception

Area of Science:

• Psychology
• Neuroscience
• Affective Computing

Background:

• Research on emotion expression (EE) perception is rapidly advancing across multiple scientific disciplines.
• Existing corpora for empirical research on EE are reviewed.
• A need exists for dynamic, multimodal emotion expression datasets.

Purpose of the Study:

• To introduce the Geneva Multimodal Emotion Portrayals Core Set (GEMEP-CS), a novel dynamic multimodal corpus.
• To outline the design features and justification for the GEMEP-CS.
• To present validation data for the core set selection and introduce an associated microcoded database.

Main Methods:

• Development of a dynamic, multimodal corpus of emotion expressions.
• Detailed validation of the core set selection through presented data.
• Creation of an associated database with microcoded facial, vocal, and body action elements and observer ratings.

Main Results:

• The article details the design and validation of the GEMEP-CS corpus.
• Validation data supporting the core set selection are presented and discussed.
• An associated database with microcoded action elements and observer ratings is introduced.

Conclusions:

• The GEMEP-CS provides a valuable new resource for studying emotion expression perception.
• The corpus and associated database facilitate multimodal research in affective science.
• This work advances the availability of dynamic, validated emotion expression datasets for empirical investigation.