A comprehensive review in affective computing: an exploration of artificial intelligence in unimodal and multimodal emotion recognition systems

• Published in: International journal of speech technology Vol. 28; no. 2; pp. 541 - 563
• Main Authors: Kapase·, Ajay Babasaheb, Uke, Nilesh
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2025; Springer Nature B.V
• Subjects: Affective computing; Anatomical systems; Artificial Intelligence; Deep learning; Emotion recognition; Emotional factors; Emotions; Engineering; Human-computer interaction; Human-computer interface; Interdisciplinary aspects; Machine learning; Multimodality; Recognition; Signal,Image and Speech Processing; Social Sciences; Deep learning; Affective computing; Multimodal emotion recognition; Unimodal emotion recognition; Machine learning
• ISSN: 1381-2416; 1572-8110
• DOI: 10.1007/s10772-025-10202-3

Affective computing is an interdisciplinary field that deals with the development of systems that can interpret, respond, and recognize human emotions. Emotion Recognition (ER) systems have a significant impact on improving the Human-Computer Interactions (HCI) by allowing devices to comprehend and respond to users’ emotional states. This review provides an extensive overview of the literature on emotion recognition, discussing the application of both physical and physiological signals for detecting emotions. This review primarily deals with unimodal and multimodal emotion recognition systems and compares different traditional methods based on machine learning, deep learning, and optimization algorithms. The primary challenges and limitations of unimodal and multimodal emotion recognition systems are also emphasized, and future research avenues are suggested at the end. It attempts to give an overall picture of the latest developments, making researchers aware of the latest techniques in this area. The research features an examination of more than 65 standard research papers on a broad spectrum of technical subjects, such as unimodal-emotion techniques and multimodal-emotion techniques.