计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (11): 1-9.DOI: 10.3778/j.issn.1002-8331.2003-0084

• 热点与综述 • 上一篇    下一篇

语音情感识别综述

孙晓虎,李洪均   

  1. 1.南通大学 信息科学技术学院,江苏 南通 226019
    2.计算机软件新技术国家重点实验室(南京大学),南京 210093
    3.南通大学 智能信息技术研究中心,江苏 南通 226019
    4.南通大学 通科微电子学院,江苏 南通 226019
  • 出版日期:2020-06-01 发布日期:2020-06-01

Overview of Speech Emotion Recognition

 SUN Xiaohu, LI Hongjun   

  1. 1.School of Information Science and Technology, Nantong University, Nantong, Jiangsu 226019, China
    2.State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210093, China
    3.Research Center for Intelligent Information Technology, Nantong University, Nantong, Jiangsu 226019, China
    4.TONGKE School of Microelectronics, Nantong University, Nantong, Jiangsu 226019, China
  • Online:2020-06-01 Published:2020-06-01

摘要:

人类的语音情感变化是一个抽象的动态过程,难以使用静态信息对其情感交互进行描述,而人工智能的兴起为语音情感识别的发展带来了新的契机。从语音情感识别的概念和在国内外发展的历史进程入手,分别从5个方面对近些年关于语音情感识别的研究成果进行了归纳总结。介绍了语音情感特征,归纳总结了各种语音特征参数对语音情感识别的意义。分别对语音情感数据库的分类及特点、语音情感识别算法的分类及优缺点、语音情感识别的应用以及语音情感识别现阶段所遇到的挑战进行了详细的阐述。立足于研究现状对语音情感识别的未来研究及其发展进行了展望。

关键词: 语音情感识别, 情感特征, 情感识别算法, 人工智能

Abstract:

The change of human speech emotion is an abstract dynamic process, and it is difficult to describe its emotional interaction using static information. The rise of artificial intelligence has brought new opportunities for the development of speech emotion recognition. Starting from the concept of speech emotion recognition and the historical process of development at home and abroad, the research results of speech emotion recognition in recent years are summarized from five aspects. Firstly, the speech emotion characteristics are introduced, and the significance of various speech characteristic parameters on speech emotion recognition is summarized. Secondly, the classification and characteristics of speech emotion database, the classification and advantages and disadvantages of speech emotion recognition algorithms, the application of speech emotion recognition, and the challenges encountered in speech emotion recognition are described in detail. Finally, based on the current research situation, the future research and development of speech emotion recognition are prospected.

Key words: speech emotion recognition, emotional characteristics, emotion recognition algorithm, artificial intelligence