The Applications of Facial Expression Recognition in Human-computer Interaction

Abstract

Facial expression is an important part of non-verbal communication and one common means of human communication. Expression recognition, as one of the important development directions of human-computer interaction, can improve the fluency, accuracy and naturalness of interaction. In recent years, deep learning using feature extraction of image data based on convolutional neural networks (CNN) has become more and more popular. Their popularity stems from their ability to extract good features from image data, for DCNN's computationally intensive tasks can be run on the GPU to achieve high performance at very low consumption. This algorithm can achieve much higher accuracy than traditional ones, making it possible to commercialize and utilize. This paper introduces the basic principles and methods of expression recognition, and classifies the common recognition methods. The cases of improving recognition rate and robustness after applying multiple algorithms are reviewed in detail. And the matters to be solved for further application of this method are also discussed. Additionally, the technical methods and approaches of expression recognition in industrial design, particularly in emotional interaction design of industrial products, are elucidated.