To Improve Is to Change: Towards Improving Mood Prediction by Learning Changes in Emotion

TL;DR

This paper introduces a framework for mood prediction that incorporates emotional change information, demonstrating that modeling both mood and emotion change enhances affective state recognition accuracy.

Contribution

It is the first to model the interplay between mood and emotional change in computational mood prediction using multimodal neural networks.

Findings

Incorporating emotional change labels improves mood prediction accuracy.

Multimodal models outperform unimodal models in mood recognition.

Modeling emotion and mood together enhances affective state analysis.

Abstract

Although the terms mood and emotion are closely related and often used interchangeably, they are distinguished based on their duration, intensity and attribution. To date, hardly any computational models have (a) examined mood recognition, and (b) modelled the interplay between mood and emotional state in their analysis. In this paper, as a first step towards mood prediction, we propose a framework that utilises both dominant emotion (or mood) labels, and emotional change labels on the AFEW-VA database. Experiments evaluating unimodal (trained only using mood labels) and multimodal (trained with both mood and emotion change labels) convolutional neural networks confirm that incorporating emotional change information in the network training process can significantly improve the mood prediction performance, thus highlighting the importance of modelling emotion and mood simultaneously for improved performance in affective state recognition.