The Many Faces of Anger: A Multicultural Video Dataset of Negative Emotions in the Wild (MFA-Wild)
This work addresses the need for culturally nuanced emotion datasets for researchers in affective computing and AI, though it is incremental as it builds on existing emotion dataset efforts.
The authors tackled the problem of cultural variation in anger expression by curating the first in-the-wild multicultural video dataset, using culture-fluent annotators to label videos with 6 labels and 13 emojis, and provided a baseline multi-label classifier to demonstrate emojis as a language-agnostic annotation tool.
The portrayal of negative emotions such as anger can vary widely between cultures and contexts, depending on the acceptability of expressing full-blown emotions rather than suppression to maintain harmony. The majority of emotional datasets collect data under the broad label ``anger", but social signals can range from annoyed, contemptuous, angry, furious, hateful, and more. In this work, we curated the first in-the-wild multicultural video dataset of emotions, and deeply explored anger-related emotional expressions by asking culture-fluent annotators to label the videos with 6 labels and 13 emojis in a multi-label framework. We provide a baseline multi-label classifier on our dataset, and show how emojis can be effectively used as a language-agnostic tool for annotation.