OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition
Electroencephalogram(EEG) signals are generated from central nervous system which are difficult to disguise, leading to its popularity in emotion recognition. Recently,semi-supervisedlearning exhibits promisingemotion recognition performance by involving unlabeled EEG data into model training. Howev...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9775684/ |
| _version_ | 1797805139454590976 |
|---|---|
| author | Yong Peng Fengzhe Jin Wanzeng Kong Feiping Nie Bao-Liang Lu Andrzej Cichocki |
| author_facet | Yong Peng Fengzhe Jin Wanzeng Kong Feiping Nie Bao-Liang Lu Andrzej Cichocki |
| author_sort | Yong Peng |
| collection | DOAJ |
| description | Electroencephalogram(EEG) signals are generated from central nervous system which are difficult to disguise, leading to its popularity in emotion recognition. Recently,semi-supervisedlearning exhibits promisingemotion recognition performance by involving unlabeled EEG data into model training. However, if we first build a graph to characterize the sample similarities and then perform label propagation on this graph, these two steps cannotwell collaborate with each other. In this paper, we propose an OptimalGraph coupledSemi-Supervised Learning (OGSSL) model for EEG emotion recognition by unifying the adaptive graph learning and emotion recognition into a single objective. Besides, we improve the label indicator matrix of unlabeledsamples in order to directly obtain theiremotional states. Moreover, the key EEG frequency bands and brain regions in emotion expression are automatically recognized by the projectionmatrix of OGSSL. Experimental results on the SEED-IV data set demonstrate that 1) OGSSL achieves excellent average accuracies of 76.51%, 77.08% and 81.29% in three cross-sessionemotion recognition tasks, 2) OGSSL is competent for discriminative EEG feature selection in emotion recognition, and 3) the Gamma frequency band, the left/righttemporal, prefrontal,and (central) parietal lobes are identified to be more correlated with the occurrence of emotions. |
| first_indexed | 2024-03-13T05:47:32Z |
| format | Article |
| id | doaj.art-25f16afacbec4a0996c247d1173fd58b |
| institution | Directory Open Access Journal |
| issn | 1558-0210 |
| language | English |
| last_indexed | 2024-03-13T05:47:32Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| spelling | doaj.art-25f16afacbec4a0996c247d1173fd58b2023-06-13T20:07:14ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1558-02102022-01-01301288129710.1109/TNSRE.2022.31754649775684OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion RecognitionYong Peng0https://orcid.org/0000-0003-1208-972XFengzhe Jin1https://orcid.org/0000-0003-3841-6573Wanzeng Kong2https://orcid.org/0000-0002-0113-6968Feiping Nie3https://orcid.org/0000-0002-0871-6519Bao-Liang Lu4https://orcid.org/0000-0001-8359-0058Andrzej Cichocki5https://orcid.org/0000-0002-8364-7226School of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaSchool of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaSchool of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaSchool of Artificial Intelligence, OPtics and ElectroNics (iOPEN), Northwestern Polytechnical University, Xi’an, ChinaDepartment of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, ChinaCenter for Computational and Data-Intensive Science and Engineering, Skolkovo Institute of Science and Technology, Moscow, RussiaElectroencephalogram(EEG) signals are generated from central nervous system which are difficult to disguise, leading to its popularity in emotion recognition. Recently,semi-supervisedlearning exhibits promisingemotion recognition performance by involving unlabeled EEG data into model training. However, if we first build a graph to characterize the sample similarities and then perform label propagation on this graph, these two steps cannotwell collaborate with each other. In this paper, we propose an OptimalGraph coupledSemi-Supervised Learning (OGSSL) model for EEG emotion recognition by unifying the adaptive graph learning and emotion recognition into a single objective. Besides, we improve the label indicator matrix of unlabeledsamples in order to directly obtain theiremotional states. Moreover, the key EEG frequency bands and brain regions in emotion expression are automatically recognized by the projectionmatrix of OGSSL. Experimental results on the SEED-IV data set demonstrate that 1) OGSSL achieves excellent average accuracies of 76.51%, 77.08% and 81.29% in three cross-sessionemotion recognition tasks, 2) OGSSL is competent for discriminative EEG feature selection in emotion recognition, and 3) the Gamma frequency band, the left/righttemporal, prefrontal,and (central) parietal lobes are identified to be more correlated with the occurrence of emotions.https://ieeexplore.ieee.org/document/9775684/Electroencephalogram (EEG)emotion recognitionfeature selectiongraph learningsemi-supervised learning |
| spellingShingle | Yong Peng Fengzhe Jin Wanzeng Kong Feiping Nie Bao-Liang Lu Andrzej Cichocki OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition IEEE Transactions on Neural Systems and Rehabilitation Engineering Electroencephalogram (EEG) emotion recognition feature selection graph learning semi-supervised learning |
| title | OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition |
| title_full | OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition |
| title_fullStr | OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition |
| title_full_unstemmed | OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition |
| title_short | OGSSL: A Semi-Supervised Classification Model Coupled With Optimal Graph Learning for EEG Emotion Recognition |
| title_sort | ogssl a semi supervised classification model coupled with optimal graph learning for eeg emotion recognition |
| topic | Electroencephalogram (EEG) emotion recognition feature selection graph learning semi-supervised learning |
| url | https://ieeexplore.ieee.org/document/9775684/ |
| work_keys_str_mv | AT yongpeng ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition AT fengzhejin ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition AT wanzengkong ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition AT feipingnie ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition AT baolianglu ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition AT andrzejcichocki ogsslasemisupervisedclassificationmodelcoupledwithoptimalgraphlearningforeegemotionrecognition |