Building a three-level multimodal emotion recognition framework

Abstract Multimodal emotion detection has been one of the main lines research in field Affective Computing (AC) recent years. detectors aggregate information coming from different channels or modalities to determine what users are expressing with a higher degree accuracy. However, despite benefits offered by this kind detectors, their presence real implementations is still scarce for various reasons. In paper, we propose technology-agnostic framework, HERA, facilitate creation multimodal offering tool characterized its modularity and interface-based programming approach adopted development. HERA (Heterogeneous Emotional Results Aggregator) offers an architecture integrate services heterogeneous results produce final result using common format. This proposal constitutes step forward development providing manage fuse produced them sensible way. We assessed validity testing system several developers no previous knowledge about affective technology detection. The assessment was performed applying Computer System Usability Questionnaire Twelve Cognitive Dimensions Questionnaire, used Visual Studio group at Microsoft, obtaining positive important feedback future versions system.