The Many Faces of Anger: A Multicultural Video Dataset of Negative Emotions in the Wild (MFA-Wild)

TL;DR

This paper introduces MFA-Wild, a multicultural in-the-wild video dataset capturing diverse negative emotional expressions like anger, with multi-label annotations including emojis, and provides baseline classification results highlighting emoji utility for annotation.

Contribution

The work presents the first multicultural in-the-wild video dataset of negative emotions with multi-label annotations and explores emoji-based labeling as a language-agnostic approach.

Findings

Baseline multi-label classifier performance demonstrated.

Emojis effectively used as annotation tools.

Dataset captures cultural variability in anger expressions.

Abstract

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.