Deep Learning based Facial Expression Recognition for Psychological Health Analysis

Facial Expression Recognition is known for its efficiency and its stimulating job in this automated world. Facial Expressions are the easiest way for human being to express their feelings. Facial expression plays a major role in communicating non-verbally. This paper summarizes the Facial Expression Recognition (FER) techniques based on deep learning. FER technique’s performance is compared based on the amount of expressions recognized and the difficulty of algorithms in CNN. FER 2013 database is been used here. Recently, the CNN (Convolutional Neural Networks) has gained the reputation within the field of deep learning owing to their effective design and also the ability to produce smart results without manual feature extraction from the raw information. This paper investigates the effectiveness of CNN with Radial Basis Function for expression recognition. The experimental results shows that the proposed method provide relatively better accuracy for FER 2013 dataset.