摘要情感在人类的感知，决策等过程扮演着重要的角色。长期以来情感智能研究只存在于心理学和认知科学领域，随着人工智能的发展，情感智能跟计算机技术结合产生了情感计算，这就为情感识别提供了更广阔的空间。语言是人类交际最重要的交流工具，人类的语言中不仅包含了文字符号信息，而且还包含了人们的感情和情绪等信息。例如，同一句话，往往由于说话者的情感不同，其含义及给听者的印象就会不同，从这点看，我们认为通过分析语音来判断人的情感是可能的。如何从语音中自动识别说话者的情感状态近年来受到各领域研究者的广泛关注，综合起来研究的问题主要包括情感理论模型，语音资料的选择，情感特征参数以及情感识别方法。60593

关键词 共振峰提取 基音提取 语音 情感识别

毕业论文毕业设计说明书（论文）外文摘要

Title    Mandarin emotional state of surprise and  sadness emotion feature extraction                                                 

Abstract

Emotion plays an important role in human perception, decision-making processes. Emotional intelligence has long been studied only in the field of psychology and cognitive science,but  with the development of artificial intelligence, emotional intelligence with computer technology combine to produce the affective computing, which provides emotion recognition a broader space. Language is the most important tools in human communication, human language includes not only the text symbol information, but also includes people's feelings, emotions and other information. For example, the same sentence, as the speaker's emotions often different, and their meaning will be different to the listener the impression that, from this point of view, we believe that the analysis to determine the emotional speech is possible.How to automatically recognize the emotional state from the speaker's voice noticed researchers in various fields in recent years, taken together, research issues include emotional theoretical models, the choice of voice data, emotional characteristic parameters and emotional recognition.

Keywords  Formant extraction Pitch extraction Speech emotion recognition

1 绪论 1

1.1  研究背景及意义 1

1.2  语音情感发展识别现状 1

1.2.1  情感的定义及分类 1

1.2.2  声学基础 2

1.2.3  语音信号数字化和预处理 2

1.2.4  语音信号时域分析 3

1.2.5  语音信号频域分析 4

1.2.6  语音信号情感特征参数分析及提取 6

1.2.7  语音情感识别主要方法 8

1.3  本文主要工作 10

2  语音信号情感辨识 11

2.1  情感特征参数选取 11

2.1.1  基音频率参数 11

2.1.2  共振峰参数 12

2.2  情感特征参数获取 12

2.2.1  基频参数提取 12

2.2.2  共振峰参数提取 15

2.3  辨识方法