A STUDY ON THE DIFFERENT TECHNIQUES FOR FACIAL EMOTION IDENTIFICATION AND CLASSIFICATION

Facial expression plays a significant role in our everyday lives. These expressions help in identifying the frame of mind of people. Henceforth, it is used in various fields like medical psychology, criminal investigations, job recruitment, etc., to study human behaviour and analyze people. In the emerging field of artificial intelligence, various categories and analyses are developed to identify and classify facial emotions. Various methods have been implemented, from image analysis to live facial emotion recognition. In our proposed work, we have implemented micro-expression based real-time facial emotion recognition (FER) using a convolution neural network in conjunction with the artificial bee colony algorithm (ABC). Datasets like FER2013, Google Data Set, CK+ and in the wild dataset were used for the training process. We achieved a 95% to 99% accuracy in real-time video-based facial recognition.