Design and Implementation of Face Emotion Recognition System Based on CNN Mini_Xception Frameworks

Abstract

In the new era, the rapid development of information technology makes the face recognition technology in artificial intelligence also develop rapidly, among which, face expression recognition has become a research hotspot. In recent years, due to deep learning, convolutional neural network and multilayer perceptron and other related algorithms have become the research focus of scholars, so their wide application in the field of facial expression recognition is also the direction of face emotion recognition exploration and research. In addition, computer software is widely used in human daily life, so it is also very important to design and implement an intelligent, real-time and universally applicable UI interface for facial emotion recognition system. Therefore, this paper firstly completed the facial expression recognition model based on the Mini_Xception framework of CNN by training the FER2013 expression database. Secondly, the system UI interface is designed and realized through PyQT5, OpenCV, Keras and other libraries. The final results show that the system is based on the algorithm model, and the UI interface designed and realized can not only recognize the saved pictures, but also recognize the real-time face emotions through the camera, and the overall effect of the system is outstanding.