Development of Facial Emotion Recognition using Convolutional Neural Network
Emotion recognition is a field in image processing that has a wide usage in fields such as robotics engineering, educational purposes and security management. There are various approaches in solving the FER (Facial Expression Recognition) problem. They can be categorized into A) Static single images and B) Image sequences. Different techniques such as the Multi-layer perceptron Model, the k-nearest neighbor and SVM (Support Vector Machines) were used for solving the FER. The methods were used to extract features like the local binary patterns, face-landmark features, and also texture features. Neural Networks have gained popularity among all these methods due to which they are extensively used for FER. Recently, CNN (Convolutional Neural Networks) has been in use in the deep learning field due to its potential to come up with good results without the need of extracting features manually. The various facial expression recognition techniques based related to CNN are used and shows steps that are needed or required for the using CNN in FER. The paper includes an in-depth analysis of the CNN based approaches and the issues that occur in the FER while choosing CNN.
Keywords: Convolutional Neural Network (CNN), Facial Expression Recognition (FER, Image Preprocessing, Classification Algorithm.