ColEmo: A Flexible Open Source Software Interface for Collecting Emotion Data

Abstract

Data collection is a critical challenge in Emotion Recognition (ER), especially as demand grows for in-the-wild data that includes contextual information. The collection of data is continuously needed to investigate new ER modalities, sensors, stimuli, models, and methods. This necessitates the development of tools and frameworks that facilitate emotion data collection. In this paper, we introduce ColEmo, an open-source software interface for collecting emotion data. ColEmo was developed using Flutter allowing it to be compiled for both desktop and mobile devices. The architecture and interface of ColEmo provides a high degree of flexibility to be customized or extended to suit specific experiment requirements. We tested ColEmo in an ER data collection study which was extended to include Voice Activity Detection (VAD) and motion context, demonstrating its effectiveness in lab environments. Further investigation is recommended to evaluate ColEmo's potential for in-the-wild data collection setups.