Description: 本实验语音库为免费的柏林情感语音库，其采样频率为16KHZ，16bit量化。该语音库共有500 句情感语音信号，分别由十名专业演员(5 男，5 女)在不同情感状态下（高兴、愤怒、平静、悲伤、害怕、厌烦、憎恨）朗读十句不同文本的德语组成。本实验选取其中的部分情感（高兴、愤怒、悲伤）加以识别。仿真实验环境为MATLAB7.0。
实验选取的情感特征为短时平均能量、短时平均幅度、基频和短时过零率。为了降低不同人在表达不同情感时的个人差异造成的影响，本文实验过程中将提取的情感特征进行归一化处理。归一化采取将同一个人的三种情感语音信号的情感特征放在一起归一化处理，并将归一化后的情感特征作为KNN分类器的训练样本和测试样本。
该实验使用了传统的KNN算法和改进算法两种识别方式，通过实验结果对比了各自的特点。
-This experiment YuYinKu for free Berlin YuYinKu emotion, the sampling frequency for 16 KHZ, 16 bit quantification. The YuYinKu 500 words of emotional speech signal, respectively by ten professional actors (5 male, five female) in different emotional state (happy, anger, sadness, fear, calm, bored, hate) read ten sentences in different text of the German. This experiment selection of part of the emotions (joy, anger, sadness) is recognized. The simulation experiment environment for MATLAB7.0.
The selection of emotional characteristics for short-term average energy, short-term average amplitude, frequency, and the short zero rate. In order to reduce the different people in different emotional expression of the impact of individual differences, the experiment process will extract emotional characteristics is normalized. The normalized will take the same three kind of emotional speech signal emotional features normalized put together, and after the normalized the emotional characteristics Platform: |
Size: 6241280 |
Author:相容 |
Hits:
Description: 基于特定人情感的语音识别，语音情感特征提取；语音情感分类器的设计，完成了一个特定人语音情感识别的初步系统。内容很丰富-In this article, we have accomplished a system for emotion recognition of speech by establishing a database of speech with certain emotion by certain people analyzing emotion features and designing a sort system for emotion recognition Platform: |
Size: 259072 |
Author:liuna |
Hits:
Description: 基于LPMCC的语音识别系统实现
语音识别可实现人机交互和语音控制，在X-业控制、消费电子等领域都有广泛应用。结合人发音的生理
结构的特点，使用LPMCC(LPC倒谱美尔变换)作为特征向量，采用动态规划算法作为核心识别算法，在TMS320VC5402芯片上实现了特定人、孤立词的高性能实时识别系统。-Speech Recognition System Based On LPMCC
Speech recognition can achieve human—computer interaction and voice control， and it is used in indUStrial control，consumer electronics and many other fields．Combined with the mechanism of the production of
human s voice． a high perform speech recognition system is built for specific people on TMSVC5402 DSP by using the LPMCC as the feature vector and using dynamic time warping as the matching algorithm． Platform: |
Size: 280576 |
Author:edwin |
Hits:
Description: 一款通用的语音情感识别软件LilyEmotion 1.0。 首先给出这款软件的下载链接： 包含在语音信号中的情感信息是一种很重要的信息资源，它是人们感知事物的必不可少的部分信息。语音之所以能够表达情感,是因为其中包含能体现情感特征的参数。研究认为,某种特定的情感状态所引起的语音参数变化在不同的人之间是大致相同的,仅有微小差别。因而，情感的变化能够通过语音的特征参数来反映。本软件首先提取能够表示语音情感的特征，之后对其进行识别，从而最终确定语音的感情 本软件可以识别12种语音感情，分别为： amusement anxiety cold_anger despair elation hot_anger interest other panic_fear pleasure pride relief sadness 等。这样一款软件有着非常广泛的应用前景：（1）用于自动远程电话服务中心，及时发现客户的不满情绪；（2）用于远程教学和婴儿教育，及时识别学生的情绪并做出适当的处理，从而提高教学质量；（3）也可以用于刑事侦察中自动检测犯罪嫌疑人的心理状态以及辅助测谎等。-A universal speech emotion recognition software LilyEmotion 1.0.Firstly, this software download link: emotional information contained in the speech signal is a very important information resource, it is the essential part of the information that people perceive things. The voice was able to express emotion, because it contains parameters that reflect the emotional characteristics. The studies suggest that certain emotional states caused by changes in the speech parameters are substantially the same between different people, with only minor differences. Accordingly, emotional changes can be reflected by the characteristic parameters of the speech. The software first extract able to represent the characteristics of speech emotion, after its identification, which ultimately determine the voice the feelings of the software can identify 12 kinds of voice feelings, are as follows: amusement the anxiety cold_anger flushed with fever Embarrassed by the crowd elation hot_anger interest other pa Platform: |
Size: 8072192 |
Author:Shi |
Hits:
Description: This paper is based on emotion recognition of speech signals. Recognizing person s emotional state is an important task. Platform: |
Size: 231424 |
Author:Jibz |
Hits:
Description: 语音情感识别分类，在中科大录制的语音情感数据库CASIA中来实现的(Speech emotion recognition and classification is implemented in CASIA, a speech emotion database recorded by China University of science and technology.) Platform: |
Size: 1024 |
Author:chestnut_ |
Hits:
Search - emotion recognition by speech