Abstract: ‍ ‍Speech emotion recognition is a kind of technology that uses computers to create the relationship between speech and emotion measurement， and provides computers with the ability to recognize and understand human emotions. Therefore， speech emotion recognition plays an important role in human-computer interaction and is a promising development direction in the field of artificial intelligence. Starting from the development history of speech emotion recognition and a series of conferences and competitions at home and abroad， this paper reviews the current research status of speech emotion recognition from six aspects. Firstly， the discrete and dimensional models for emotional representation are described. Secondly， the current commonly used speech emotion databases are summarized in detail. Thirdly， representative speech emotion recognition development history is reviewed in the past 20 years， and the speech emotion recognition technology based on hand-crafted speech emotion features and end-to-end framework are described， respectively. Then， the performance of speech emotion recognition in recent years is summarized， especially the major conferences and journals in the speech signal field in the past two years. Then， the applications of speech emotion recognition in driving， intelligent interaction， medical health， safety and other fields are introduced. Finally， the challenges and trends in the field of speech emotion recognition are described from the three aspects， including speech emotion database， speech emotion features， and algorithms/models. This paper aims to analyze the related work of speech emotion recognition in detail and provides a valuable reference for researchers who are engaged in speech emotion recognition research filed.