A Conv2D Based Facial Expression Recognition and a Model Thereof

Facial Expression recognition (FER) is an emerging field in Deep Learning. It is widely used in Medical, Security, online shopping, and many more. The future actions of an individual can be predicted from his/her expressions. It can improve human-computer interaction. It is a process of detecting face expressions of an individual from videos or photographs by using the features of faces to distinguish between different expressions. An approach based on Deep Learning (DL) is proposed in this work. Here Conv2D model with four layers is used to detect the facial expressions of humans. For extracting the best features from input, Max-pooling is used. The input images are converted into grayscale before training. The proposed model achieved an accuracy of 83.23%