Affective computing and artificial intelligence both rely heavily on facial expression recognition. Although human face expressions vary so often that existing techniques rely heavily on feature expression to recognize them effectively. Recently, Facial Expression Recognition (FER) has attracted a lot of interest as facial expressions are taken into account. as the informational medium with the quickest means of communication. Facial expression analysis, which is currently popular, uses deep learning techniques and provides a better comprehension of a person's thoughts or viewpoints. Compared to conventional state-of-the-art systems, the accuracy rate has significantly increased. Deep learning focuses on the stimulation of the organizational structure of human brain nerves that incorporate low-level characteristics and is currently a popular issue in the field of machine learning. In this study, we concentrate on using a conventional neural network to develop a facial expression recognition system that aids in the identification of deeper feature representations of facial expressions and therefore achieves automatic recognition. This article explores the most recent and current reviews in FER employing Convolution Neural Network (CNN) algorithms and provides a quick summary of the various FER fields of application and publically accessible databases used in FER.