A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN
Emotion recognition based on electroencephalograms has become an active research area. Yet, identifying emotions using only brainwaves is still very challenging, especially the subject-independent task. Numerous studies have tried to propose methods to recognize emotions, including machine learning...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/5/1678 |
| _version_ | 1797416632153276416 |
|---|---|
| author | Panayu Keelawat Nattapong Thammasan Masayuki Numao Boonserm Kijsirikul |
| author_facet | Panayu Keelawat Nattapong Thammasan Masayuki Numao Boonserm Kijsirikul |
| author_sort | Panayu Keelawat |
| collection | DOAJ |
| description | Emotion recognition based on electroencephalograms has become an active research area. Yet, identifying emotions using only brainwaves is still very challenging, especially the subject-independent task. Numerous studies have tried to propose methods to recognize emotions, including machine learning techniques like convolutional neural network (CNN). Since CNN has shown its potential in generalization to unseen subjects, manipulating CNN hyperparameters like the window size and electrode order might be beneficial. To our knowledge, this is the first work that extensively observed the parameter selection effect on the CNN. The temporal information in distinct window sizes was found to significantly affect the recognition performance, and CNN was found to be more responsive to changing window sizes than the support vector machine. Classifying the arousal achieved the best performance with a window size of ten seconds, obtaining 56.85% accuracy and a Matthews correlation coefficient (MCC) of 0.1369. Valence recognition had the best performance with a window length of eight seconds at 73.34% accuracy and an MCC value of 0.4669. Spatial information from varying the electrode orders had a small effect on the classification. Overall, valence results had a much more superior performance than arousal results, which were, perhaps, influenced by features related to brain activity asymmetry between the left and right hemispheres. |
| first_indexed | 2024-03-09T06:05:59Z |
| format | Article |
| id | doaj.art-16f2f1de4f66485280126c1b1c9b99ae |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T06:05:59Z |
| publishDate | 2021-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-16f2f1de4f66485280126c1b1c9b99ae2023-12-03T12:02:49ZengMDPI AGSensors1424-82202021-03-01215167810.3390/s21051678A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNNPanayu Keelawat0Nattapong Thammasan1Masayuki Numao2Boonserm Kijsirikul3Department of Computer Science and Engineering, University of California San Diego, La Jolla, CA 92093-0404, USAHuman Media Interaction, Faculty of Electrical Engineering, Mathematics and Computer Science, University of Twente, 7522 NB Enschede, The NetherlandsThe Institute of Scientific and Industrial Research, Osaka University, Mihogaoka, Ibaraki, Osaka 567-0047, JapanDepartment of Computer Engineering, Chulalongkorn University, Pathum Wan, Bangkok 10330, ThailandEmotion recognition based on electroencephalograms has become an active research area. Yet, identifying emotions using only brainwaves is still very challenging, especially the subject-independent task. Numerous studies have tried to propose methods to recognize emotions, including machine learning techniques like convolutional neural network (CNN). Since CNN has shown its potential in generalization to unseen subjects, manipulating CNN hyperparameters like the window size and electrode order might be beneficial. To our knowledge, this is the first work that extensively observed the parameter selection effect on the CNN. The temporal information in distinct window sizes was found to significantly affect the recognition performance, and CNN was found to be more responsive to changing window sizes than the support vector machine. Classifying the arousal achieved the best performance with a window size of ten seconds, obtaining 56.85% accuracy and a Matthews correlation coefficient (MCC) of 0.1369. Valence recognition had the best performance with a window length of eight seconds at 73.34% accuracy and an MCC value of 0.4669. Spatial information from varying the electrode orders had a small effect on the classification. Overall, valence results had a much more superior performance than arousal results, which were, perhaps, influenced by features related to brain activity asymmetry between the left and right hemispheres.https://www.mdpi.com/1424-8220/21/5/1678emotion recognitionEEGmachine learningCNNspatiotemporal databrainwave |
| spellingShingle | Panayu Keelawat Nattapong Thammasan Masayuki Numao Boonserm Kijsirikul A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN Sensors emotion recognition EEG machine learning CNN spatiotemporal data brainwave |
| title | A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN |
| title_full | A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN |
| title_fullStr | A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN |
| title_full_unstemmed | A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN |
| title_short | A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN |
| title_sort | comparative study of window size and channel arrangement on eeg emotion recognition using deep cnn |
| topic | emotion recognition EEG machine learning CNN spatiotemporal data brainwave |
| url | https://www.mdpi.com/1424-8220/21/5/1678 |
| work_keys_str_mv | AT panayukeelawat acomparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT nattapongthammasan acomparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT masayukinumao acomparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT boonsermkijsirikul acomparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT panayukeelawat comparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT nattapongthammasan comparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT masayukinumao comparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn AT boonsermkijsirikul comparativestudyofwindowsizeandchannelarrangementoneegemotionrecognitionusingdeepcnn |