A Short Review on Multimodal Emotion Recognition Methods

The recognition and identification of human emotion states through the combination of several signals, such as text, audio, and facial cues, is known as multimodal emotion recognition (MER).MER aims to identify human emotional states by integrating information from multiple sources, such as speech, facial expressions, text, and physiological signals.In the field of human-computer interaction (HCI), healthcare, and intelligent systems, MER is essential.With advancements in machine learning (ML), deep learning (DL), and the increasing availability of rich multimodal datasets, MER has experienced significant progress.Due to the recent advancements in deep learning (DL) methods and the growing availability of multimodal datasets, the area of MER has seen remarkable development, leading to several important research achievements.There are, however, very few in-depth and targeted evaluations of ML and DL-based MER methods.This survey seeks to offer a review of MER depending on ML and DL techniques.A review that considers 15 articles based on MER is considered in this survey.The review includes a study on miscellaneous aspects.The methods including DL-oriented and ML-oriented approaches for MER are examined in this work.Moreover, the types of multi-modalities taken into account in each of the articles are reviewed.Examination of varied performances and consequently the best attainments are reviewed.Finally, the research gaps on MER are conferred.