A systematic analysis of machine learning algorithms for human emotion detection using facial expression

Abstract

Nonverbal signs, which are extremely significant in human interaction, are conveyed through facial gestures. A significant feature in normal human-machine interfaces could be automated interpretation of facial expressions; it may also be used in primary care and social psychology. Due to the fact that humans immediately perceive facial movements for all intensive purposes, recognition of expressions by computer is indeed a task. A face expression may involve deformities of facial sections and their spatial relationships, along with changes in skin pigmentation, from the standpoint of automatic identification. When it comes to facial expressions, this study is dedicated to refining the process of identification of seven specific emotions (joy, sorrow, terror, rage, surprise, disgust and neutrality). The established approaches to the construction of emotion detection systems were examined centered on expressions of human face. This study assesses experimental studies and analytical articles for emotion recognition, as well as a variety of other approaches that have been applied or investigated. The findings of this study provide specific procedures for each category of expression and its severity, as well as a grading system for them. This article can be informative who use facial emotion assessment and interpretation and have to select a tool that is appropriate for their needs or make alternate choices.