A survey on the development of intelligent robots in speech emotion recognition

Abstract

Speech emotion recognition, an important branch of affective computing, has attracted much attention recently, which is of great significance for the realization of natural and harmonious human-robot interaction. Many researches have been carried out and remarkable results have been achieved in this field. At the same time, there are still many problems to be tackled urgently. Therefore, it is necessary to summarize the previous works and find out the problems in speech emotion recognition, so as to provide guidance for further research. This paper systematically summarizes the general process of speech emotion recognition, including speech emotional databases and various leading emotion classification models. By listing some outstanding works of speech emotion recognition in recent 20 years, this paper compares and analyzes the highlights and shortcomings of these works. It can be seen that people show more interest in deep learning in which features are mostly extracted automatically than the traditional machine learning methods. Finally, the main problems in the field of speech emotion recognition and the direction of further exploration are summarized in order to promote speech emotion recognition to a new stage.