Emotion Classification using Feature Extraction of Facial Expression

Automatic Emotion recognition based on human facial expression is an intriguing research topic that has been presented and used in a diverse range of fields, which include personal security, general wellbeing, and human-machine interactions. Researchers in this field are very much curious in developing the strategies for extracting the facial expressions to predict the emotion to develop applications under HCI. This research aims to provide an overview of recent deep learning-based emotion recognition detection research (FER). We showcase the contributions explained, the design and databases used, and demonstrate the advances made by contrasting the significant indication and the data produced. The proposed work is examined on three benchmark datasets, FER2013, JAFFE, FERG, etc., and compared with the live video to check the performance. The existing state-of-art methods also were compared for further evaluation. The proposed system shows better performance with an accuracy of 98.03% for the CK+ dataset, JAFFE dataset with 90.56% and FER dataset with 87.36%.