Unimodal approaches for emotion recognition: A systematic review

Affective computing is a rising interdisciplinary field of research spanning the areas from artificial intelligence, natural language processing to cognitive and social sciences. Potential applications comprise of man-machine interaction, healthcare, entertainment, teaching, marketing and many more. Despite the increasing number of papers published in the domains of affective computing, emotion recognition, and human-computer interaction (HCI), there are still gaps in the comprehensive literature review that covers all relevant studies in a single study, which this review attempts to address. As a result, this study provides a systematic literature review (SLR) on existing modalities (unimodals) for emotion recognition, emotion models, and trends in relevant studies by selecting articles published from January 2010 to June 2021. To ensure the retrieval of all relevant studies, a review protocol is used that includes both automatic and manual searches. Based on the research questions, the final 129 papers are reviewed and relevant information is extracted. This SLR provides future research directions to assist novice researchers and practitioners in more efficiently utilizing affective computing techniques.