基于稀疏表示的脑电(EEG)情感分类
|
摘要
计算机对人类情绪与情感的识别研究已经成为了脑机接口领域的研究热点。通过分析人类在生活中的各种情感状态,提取脑电信号的特征并对情感状态进行识别、分类是情感智能化领域的重要方向。针对基于音乐视频诱导的情感数据集DEAP进行了研究,提取脑电信号的频域特征后,提出了采用加速近邻梯度(APG)算法和正交匹配(OMP)算法求解稀疏编码的稀疏表示分类模型进行情感分类,并与支持向量机(SVM)算法进行效果比较。实验结果表明,APG算法通过l1范数正则近似求解以其快速的收敛速度在情感数据集上有着较好的分类表现,而OMP算法与SVM算法的分类效果相差无几,实现了情感脑电信号的分类。
Computer recognition of human emotion has become a hot topic in the field of brain computer interface( BCI) in recently years. By analyzing the various emotional states in people's life,extracting the features of EEG and classifying emotional states is an important direction in the field of emotional intelligence. Based on the emotion data set induced by the music video,this research extracted the frequency-domain features of EEG. After that,the accelerated proximal gradient( APG) and orthogonal matching pursuit( OMP) algorithms for the sparse representation method were adopted to classify the EEG signals. By comparing with other algorithms,the experimental results show that the APG with l1 norm performs well in the emotion data set with fast convergence speed,and the greedy idea based OMP algorithm can achieve the same effect with other algorithms. The comparative analysis show the effectiveness and feasibility of the proposed method for emotional EEG signals classification.
Computer recognition of human emotion has become a hot topic in the field of brain computer interface( BCI) in recently years. By analyzing the various emotional states in people's life,extracting the features of EEG and classifying emotional states is an important direction in the field of emotional intelligence. Based on the emotion data set induced by the music video,this research extracted the frequency-domain features of EEG. After that,the accelerated proximal gradient( APG) and orthogonal matching pursuit( OMP) algorithms for the sparse representation method were adopted to classify the EEG signals. By comparing with other algorithms,the experimental results show that the APG with l1 norm performs well in the emotion data set with fast convergence speed,and the greedy idea based OMP algorithm can achieve the same effect with other algorithms. The comparative analysis show the effectiveness and feasibility of the proposed method for emotional EEG signals classification.
引文
[1] Karthick N G,Ahamed V I T,Paul J K. Music and the EEG:a studyusing nonlinear methods[C]//Proc of IEEE International Conferenceon Biomedical and Pharmaceutical Engineering. Piscataway,NJ:IEEEPress,2006:424-427.
[2] Nie Dan,Wang Xiaowei,Shi Lichen,et al. EEG-based emotion recog-nition during watching movies[C]//Proc of International IEEE/EM-BS Conference on Neural Engineering. Piscataway,NJ:IEEE Press,2011:667-670.
[3] Sander K,Christian M,Mohammad S,et al. DEAP:a database foremotion analysis using physiological signals[J]. IEEE Trans on Af-fective Computing,2011,3(1):18-31.
[4] Russell J A. A circumplex model of affect[J]. Journal of Personali-ty and Social Psychology,1980,39(6):1161-1178.
[5] Davidson R J,Jackson D C,Kalin N H. Emotion,plasticity,context,and regulation[J]. Psychological Bulletin,2000,126(6):890-909.
[6] Petrantonakis P C,Hadjileontiadis L J. Emotion recognition from EEGusing higher order crossings[J]. IEEE Trans on Information Tech-nology in Biomedicine,2010,14(2):186-197.
[7] Runyon R P,Coleman K A,Pittenger D J. Fundamentals of behavioralstatistics[J]. Stomatologie Der Ddr,1996,36(12):733.
[8]李孔震,王炳和,娄昊,等.基于小波变换和二维非负矩阵分解的人脸识别算法[J].计算机应用研究,2013,30(4):1275-1277,1280.(Li Kongzhen,Wang Binghe,Lou Hao,et al. Face recognitionalgorithm based on wavelet transform and two-dimensional non-nega-tive matrix decomposition[J]. Application Research of Compu-ters,2013,30(4):1275-1277,1280.)
[9] Olshausen B A,Field D J. Emergence of simple-cell receptive fieldproperties by learning a sparse code for natural images[J]. Nature,1996,381(6583):607-609.
[10]Tsaig Y,Donoho D L. Extensions of compressed sensing[J]. SignalProcessing,2006,86(3):549-571.
[11]刘继承,陈佳伟.基于改进St OMP算法图像压缩感知重构[J].计算机应用研究,2016,33(9):2869-2872,2877.(Liu Jicheng,Chen Jiawei. Image compression sensing reconstruction based on im-proved StOMP algorithm[J]. Application Research of Compu-ters,2016,33(9):2869-2872,2877.)
[12]Yin Jun,Liu Zhonghua,Jin Zhong,et al. Kernel sparse representationbased classification[J]. Neurocomputing,2012,77(1):120-128.
[13]Mallat S G,Zhang Zhifeng. Matching pursuits with time-frequency dic-tionaries[J]. IEEE Trans on Signal Processing,1993,41(12):3397-3415.
[14]Pati Y C,Rezaiifar R,Krishnaprasad P S. Orthogonal matching pur-suit:recursive function approximation with applications to wavelet de-composition[C]//Proc of IEEE Conference Record of the 27th Asilo-mar Conference on Signals,Systems and Computers. Piscataway,NJ:IEEE Press,2002:40-44.
[15]Donoho D L,Tsaig Y,Drori I,et al. Sparse solution of underdeter-mined systems of linear equations by stagewise orthogonal matchingpursuit[J]. IEEE Trans on Information Theory,2012,58(2):1094-1121.
[16]孙君顶,赵慧慧.图像稀疏表示及其在图像处理中的应用[J].红外技术,2014,36(7):533-537.(Sun Junding,Zhao Huihui. Imagesparse representation and its application in image processing[J]. In-frared Technology,2014,36(7):533-537.)
[17]Aronsson G,Crandall M,Juutinen P. A tour of the theory of absolutelyminimizing functions[J]. Bulletin of the American MathematicalSociety,2004,41(4):439-505.
[18]Beck A,Teboulle M. Fast gradient-based algorithms for constrained to-tal variation image denoising and deblurring problems[J]. IEEETrans on Image Processing,2009,18(11):2419-2434.
[19]黄应清,赵锴,蒋晓瑜.基于核空间类间平均距的径向基函数—支持向量机特征选择算法[J].计算机应用研究,2012,29(12):4556-4559.(Huang Yingqing,Zhao Kai,Jiang Xiaoyu. Radial basisfunction-support vector machine feature selection algorithm based onmean distance between classes in kernel space[J]. Application Re-search of Computers,2012,29(12):4556-4559.)
[20]Li Mu,Lu Baoliang. Emotion classification based on gamma-band EEG[C]//Proc of IEEE International Conference on Medicine andBiology Society. Piscataway,NJ:IEEE Press,2009:1223-1226.
[2] Nie Dan,Wang Xiaowei,Shi Lichen,et al. EEG-based emotion recog-nition during watching movies[C]//Proc of International IEEE/EM-BS Conference on Neural Engineering. Piscataway,NJ:IEEE Press,2011:667-670.
[3] Sander K,Christian M,Mohammad S,et al. DEAP:a database foremotion analysis using physiological signals[J]. IEEE Trans on Af-fective Computing,2011,3(1):18-31.
[4] Russell J A. A circumplex model of affect[J]. Journal of Personali-ty and Social Psychology,1980,39(6):1161-1178.
[5] Davidson R J,Jackson D C,Kalin N H. Emotion,plasticity,context,and regulation[J]. Psychological Bulletin,2000,126(6):890-909.
[6] Petrantonakis P C,Hadjileontiadis L J. Emotion recognition from EEGusing higher order crossings[J]. IEEE Trans on Information Tech-nology in Biomedicine,2010,14(2):186-197.
[7] Runyon R P,Coleman K A,Pittenger D J. Fundamentals of behavioralstatistics[J]. Stomatologie Der Ddr,1996,36(12):733.
[8]李孔震,王炳和,娄昊,等.基于小波变换和二维非负矩阵分解的人脸识别算法[J].计算机应用研究,2013,30(4):1275-1277,1280.(Li Kongzhen,Wang Binghe,Lou Hao,et al. Face recognitionalgorithm based on wavelet transform and two-dimensional non-nega-tive matrix decomposition[J]. Application Research of Compu-ters,2013,30(4):1275-1277,1280.)
[9] Olshausen B A,Field D J. Emergence of simple-cell receptive fieldproperties by learning a sparse code for natural images[J]. Nature,1996,381(6583):607-609.
[10]Tsaig Y,Donoho D L. Extensions of compressed sensing[J]. SignalProcessing,2006,86(3):549-571.
[11]刘继承,陈佳伟.基于改进St OMP算法图像压缩感知重构[J].计算机应用研究,2016,33(9):2869-2872,2877.(Liu Jicheng,Chen Jiawei. Image compression sensing reconstruction based on im-proved StOMP algorithm[J]. Application Research of Compu-ters,2016,33(9):2869-2872,2877.)
[12]Yin Jun,Liu Zhonghua,Jin Zhong,et al. Kernel sparse representationbased classification[J]. Neurocomputing,2012,77(1):120-128.
[13]Mallat S G,Zhang Zhifeng. Matching pursuits with time-frequency dic-tionaries[J]. IEEE Trans on Signal Processing,1993,41(12):3397-3415.
[14]Pati Y C,Rezaiifar R,Krishnaprasad P S. Orthogonal matching pur-suit:recursive function approximation with applications to wavelet de-composition[C]//Proc of IEEE Conference Record of the 27th Asilo-mar Conference on Signals,Systems and Computers. Piscataway,NJ:IEEE Press,2002:40-44.
[15]Donoho D L,Tsaig Y,Drori I,et al. Sparse solution of underdeter-mined systems of linear equations by stagewise orthogonal matchingpursuit[J]. IEEE Trans on Information Theory,2012,58(2):1094-1121.
[16]孙君顶,赵慧慧.图像稀疏表示及其在图像处理中的应用[J].红外技术,2014,36(7):533-537.(Sun Junding,Zhao Huihui. Imagesparse representation and its application in image processing[J]. In-frared Technology,2014,36(7):533-537.)
[17]Aronsson G,Crandall M,Juutinen P. A tour of the theory of absolutelyminimizing functions[J]. Bulletin of the American MathematicalSociety,2004,41(4):439-505.
[18]Beck A,Teboulle M. Fast gradient-based algorithms for constrained to-tal variation image denoising and deblurring problems[J]. IEEETrans on Image Processing,2009,18(11):2419-2434.
[19]黄应清,赵锴,蒋晓瑜.基于核空间类间平均距的径向基函数—支持向量机特征选择算法[J].计算机应用研究,2012,29(12):4556-4559.(Huang Yingqing,Zhao Kai,Jiang Xiaoyu. Radial basisfunction-support vector machine feature selection algorithm based onmean distance between classes in kernel space[J]. Application Re-search of Computers,2012,29(12):4556-4559.)
[20]Li Mu,Lu Baoliang. Emotion classification based on gamma-band EEG[C]//Proc of IEEE International Conference on Medicine andBiology Society. Piscataway,NJ:IEEE Press,2009:1223-1226.