Search

Facial expression recognition (FER) is emerging as an emerging and multifaceted field of study. The use of FER in areas such as healthcare, security, and safe driving has not only enhanced the credibility of these technologies, but also their integration into human computer interaction to achieve intelligent outcomes. Computational FER seeks to replicate the skill of humans in decoding facial expressions, providing important cues that complement spoken language and aid listeners in understanding. Likewise, FER's deep learning (DL) and Artificial Intelligence (AI) methodologies are meticulously designed, incorporating advanced modules for efficiency and real time processing. In light of this background, many investigations have looked at different aspects of FER. Although current surveys focus primarily on traditional technologies and generic methodologies for on premises servers, they overlook the large field of deep learning inspired by edge vision and AI assisted FER technologies. To fill this gap, the current study conducts a comprehensive and thorough analysis of the prevailing FER literature. It carefully surveys the operational framework of FER technologies, highlighting their basic and intermediate phases, as well as the underlying pattern structures. Furthermore, the study addresses the limitations inherent in current FER surveys. The exploration extends to the FER datasets, subjecting them to thorough examination, thus revealing the attendant challenges and pitfalls. In addition, it provides a comprehensive discussion of the various metrics used to measure the effectiveness of FER methods. [ABSTRACT FROM AUTHOR]