[1] Huang Chengwei, Wu Di, Zhang Xiaojun, Xiao Zhongzhe, et al. Cascaded projection of Gaussian mixture modelfor emotion recognition in speech and ECG signals [J]. Journal of Southeast University (English Edition), 2015, 31 (3): 320-326. [doi:10.3969/j.issn.1003-7985.2015.03.004]

Copy

Cascaded projection of Gaussian mixture modelfor emotion recognition in speech and ECG signals()

基于级联投影高斯混合模型的语音与心电情绪识别

Share：

Journal of Southeast University (English Edition)[ISSN:1003-7985/CN:32-1325/N]

Volumn:

31

Issue:

2015 3

Page:

320-326

Research Field:

Information and Communication Engineering

Publishing date:

2015-09-20

Info

Title:

Cascaded projection of Gaussian mixture modelfor emotion recognition in speech and ECG signals

基于级联投影高斯混合模型的语音与心电情绪识别

Author(s):

Huang Chengwei¹;  Wu Di¹;  Zhang Xiaojun¹;  Xiao Zhongzhe¹;  Xu Yishen¹;  Ji Jingjing¹;  Tao Zhi¹;  Zhao Li²

¹College of Physics, Optoelectronics and Energy, Soochow University, Suzhou 215006, China
²School of Information Science and Engineering, Southeast University, Nanjing 210096, China

黄程韦¹;  吴迪¹;  张晓俊¹;  肖仲喆¹;  许宜申¹;  季晶晶¹;  陶智¹;  赵力²

¹苏州大学物理与光电·能源学部, 苏州215006; ²东南大学信息科学与工程学院, 南京 210096

Keywords:

Gaussian mixture model;  emotion recognition;  sample adaptation;  emotion inducing

高斯混合模型;  情绪识别;  样本自适应;  情绪诱发

PACS:

TN912.3

DOI:

10.3969/j.issn.1003-7985.2015.03.004

Abstract:

A cascaded projection of the Gaussian mixture model algorithm is proposed. First, the marginal distribution of the Gaussian mixture model is computed for different feature dimensions, and a number of sub-classifiers are generated using the marginal distribution model. Each sub-classifier is based on different feature sets. The cascaded structure is adopted to fuse the sub-classifiers dynamically to achieve sample adaptation ability. Secondly, the effectiveness of the proposed algorithm is verified on electrocardiogram emotional signal and speech emotional signal. Emotional data including fidgetiness, happiness and sadness is collected by induction experiments. Finally, the emotion feature extraction method is discussed, including heart rate variability, the chaotic electrocardiogram feature and utterance level static feature. The emotional feature reduction methods are studied, including principle component analysis, sequential forward selection, the Fisher discriminant ratio and maximal information coefficient. The experimental results show that the proposed classification algorithm can effectively improve recognition accuracy in two different scenarios.

提出了一种基于级联投影的高斯混合模型算法.首先, 针对不同的特征维度计算高斯混合模型的边缘概率, 依据边缘概率模型构造出多个子分类器, 每个子分类器包含不同的特征组合.采用级联结构的框架对子分类器进行动态融合, 从而获得对样本的自适应能力.其次, 在心电情感信号和语音情感信号上验证了算法的有效性, 通过实验诱发手段, 采集了烦躁、喜悦、悲伤等情感数据.最后, 探讨了情感特征参数(心率变异性、心电混沌特征, 语句级静态特征等)的提取方法.研究了情感特征的降维方法, 包括主分量分析、顺序特征选择、Fisher区分度和最大信息系数等方法.实验结果显示, 所提算法能够在2种不同的场景中有效地提高情感识别的准确率.

References:

[1] Schuller B, Rigoll G, Lang M. Hidden Markov model-based speech emotion recognition [C]//IEEE International Conference on Acoustics, Speech, and Signal Processing. Hong Kong, China, 2003, 2: 401-404.
[2] Kim J, André E. Emotion recognition based on physiological changes in music listening [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(12): 2067-2083.
[3] Khan R A, Meyer A, Konik H, et al. Framework for reliable, real-time facial expression recognition for low resolution images [J]. Pattern Recognition Letters, 2013, 34(10): 1159-1168.
[4] Steidl S. Automatic classification of emotion-related user states in spontaneous children’s speech [D].Erlangen-Nuremberg, Germany: FAU Erlangen-Nuremberg, 2009.
[5] Kockmann M, Burget L, ()/Cernocký J H. Application of speaker and language identification state-of-the-art techniques for emotion recognition [J]. Speech Communication, 2011, 53(9/10): 1172-1185.
[6] Pantelopoulos A, Bourbakis N G. A survey on wearable sensor-based systems for health monitoring and prognosis [J]. IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 2010, 40(1): 1-12.
[7] Gunes H, Pantic M. Automatic, dimensional and continuous emotion recognition [J]. International Journal of Synthetic Emotions, 2010, 1(1): 68-99.
[8] McNicholas P D, Murphy T B. Model-based clustering of microarray expression data via latent Gaussian mixture models [J]. Bioinformatics, 2010, 26(21): 2705-2712.
[9] Singh R, Pal B C, Jabr R A. Statistical representation of distribution system loads using Gaussian mixture model [J]. IEEE Transactions on Power Systems, 2010, 25(1): 29-37.
[10] Chen T, Zhang J. On-line multivariate statistical monitoring of batch processes using Gaussian mixture model [J]. Computers & Chemical Engineering, 2010, 34(4): 500-507.
[11] Jian B, Vemuri B C. Robust point set registration using Gaussian mixture models [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8): 1633-1645.
[12] Kamishima Y, Inoue N, Shinoda K. Event detection in consumer videos using GMM supervectors and SVMs [J]. EURASIP Journal on Image and Video Processing, 2013, 2013: 51-59.
[13] Yu H, Huang C W, Zhao L, et al. Research of emotional electrophysiological parameters collection system and child-emotion monitoring[J]. Chinese Journal of Electronic Devices, 2010, 33(4): 516-520.(in Chinese)
[14] Grassberger P, Procaccia I. Measuring the strangeness of strange attractors[J]. Physica D, 1983, 9(1/2): 189-208.
[15] Reshef D N, Reshef Y A, Finucane H K, et al. Detecting novel associations in large data sets [J]. Science, 2011, 334(6062): 1518-1524.

Memo

Memo:

Biographies: Huang Chengwei(1984—), male, doctor, associate professor, cwhuang@suda.edu.cn.
Foundation items: The National Natural Science Foundation of China(No.61231002, 61273266, 51075068, 61271359), Doctoral Fund of Ministry of Education of China(No.20110092130004).
Citation: Huang Chengwei, Wu Di, Zhang Xiaojun, et al.Cascaded projection of Gaussian mixture model for emotion recognition in speech and ECG signals[J].Journal of Southeast University(English Edition), 2015, 31(3):320-326.[doi:10.3969/j.issn.1003-7985.2015.03.004]

Common functions