A Novel Facial Emotion Recognition Scheme Based on Graph Mining

Recently, learning human feelings through analyzing facial expressions has been acquiring immense attention. Facial expressions play a decisive role in expressing the mental status and are vital in social interactions. Recognition and analyzation of facial expressions aids in understanding one's emotions in all situations even when verbal communication fails. This paper describes the fundamental phases involved in the emotion detection process. It explores the various face emotion detection schemes like machine learning, deep learning and others employed in literature and compares the diverse emotion recognition approaches in terms of FER techniques, datasets, emotions, number of emotions and publication years. This article then reviews the several graph-based techniques used for facial emotion recognition (FER) in existing studies and explains their benefits in the FER process. Finally, this work in this paper presents the vital findings discovered during the study for achieving further improvement in the FER process and developing better approaches for superior FER performance.