EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals
Electroencephalogram (EEG)-based emotion recognition has garnered significant attention in brain–computer interface research and healthcare applications. While deep learning models have been extensively studied, most are designed for classification tasks and struggle to accurately predict continuous...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/13/1/87 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841549150567530496 |
|---|---|
| author | Hyo Jin Jon Longbin Jin Hyuntaek Jung Hyunseo Kim Eun Yi Kim |
| author_facet | Hyo Jin Jon Longbin Jin Hyuntaek Jung Hyunseo Kim Eun Yi Kim |
| author_sort | Hyo Jin Jon |
| collection | DOAJ |
| description | Electroencephalogram (EEG)-based emotion recognition has garnered significant attention in brain–computer interface research and healthcare applications. While deep learning models have been extensively studied, most are designed for classification tasks and struggle to accurately predict continuous emotional scores in regression settings. In this paper, we introduce EEG-RegNet, a novel deep neural network tailored for precise emotional score prediction across the continuous valence–arousal–dominance (VAD) space. EEG-RegNet tackles two core challenges: extracting subject-independent, emotion-relevant EEG features and mapping these features to fine-grained, continuous emotional scores. The model leverages 2D convolutional neural networks (CNNs) for spatial feature extraction and a 1D CNN for temporal dynamics, providing robust spatiotemporal modeling. A key innovation is the hybrid loss function, which integrates mean squared error (MSE) and cross-entropy (CE) with a Bernoulli penalty to enhance probability estimation and address sparsity in the emotional space. Extensive experiments on the DEAP dataset show that EEG-RegNet achieves state-of-the-art results in continuous emotional score prediction and attains 95% accuracy in fine-grained emotion classification, highlighting its scalability and precision in emotion recognition. |
| format | Article |
| id | doaj-art-796c2d66edcd4a2bbf4b65a488624cea |
| institution | Kabale University |
| issn | 2227-7390 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj-art-796c2d66edcd4a2bbf4b65a488624cea2025-01-10T13:18:12ZengMDPI AGMathematics2227-73902024-12-011318710.3390/math13010087EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG SignalsHyo Jin Jon0Longbin Jin1Hyuntaek Jung2Hyunseo Kim3Eun Yi Kim4AI&Computer Vision Laboratory, Konkuk University, Seoul 05029, Republic of KoreaAI&Computer Vision Laboratory, Konkuk University, Seoul 05029, Republic of KoreaAI&Computer Vision Laboratory, Konkuk University, Seoul 05029, Republic of KoreaAI&Computer Vision Laboratory, Konkuk University, Seoul 05029, Republic of KoreaAI&Computer Vision Laboratory, Konkuk University, Seoul 05029, Republic of KoreaElectroencephalogram (EEG)-based emotion recognition has garnered significant attention in brain–computer interface research and healthcare applications. While deep learning models have been extensively studied, most are designed for classification tasks and struggle to accurately predict continuous emotional scores in regression settings. In this paper, we introduce EEG-RegNet, a novel deep neural network tailored for precise emotional score prediction across the continuous valence–arousal–dominance (VAD) space. EEG-RegNet tackles two core challenges: extracting subject-independent, emotion-relevant EEG features and mapping these features to fine-grained, continuous emotional scores. The model leverages 2D convolutional neural networks (CNNs) for spatial feature extraction and a 1D CNN for temporal dynamics, providing robust spatiotemporal modeling. A key innovation is the hybrid loss function, which integrates mean squared error (MSE) and cross-entropy (CE) with a Bernoulli penalty to enhance probability estimation and address sparsity in the emotional space. Extensive experiments on the DEAP dataset show that EEG-RegNet achieves state-of-the-art results in continuous emotional score prediction and attains 95% accuracy in fine-grained emotion classification, highlighting its scalability and precision in emotion recognition.https://www.mdpi.com/2227-7390/13/1/87electroencephalogramemotion recognitionspatiotemporal feature extractionregressive emotion recognition |
| spellingShingle | Hyo Jin Jon Longbin Jin Hyuntaek Jung Hyunseo Kim Eun Yi Kim EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals Mathematics electroencephalogram emotion recognition spatiotemporal feature extraction regressive emotion recognition |
| title | EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals |
| title_full | EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals |
| title_fullStr | EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals |
| title_full_unstemmed | EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals |
| title_short | EEG-RegNet: Regressive Emotion Recognition in Continuous VAD Space Using EEG Signals |
| title_sort | eeg regnet regressive emotion recognition in continuous vad space using eeg signals |
| topic | electroencephalogram emotion recognition spatiotemporal feature extraction regressive emotion recognition |
| url | https://www.mdpi.com/2227-7390/13/1/87 |
| work_keys_str_mv | AT hyojinjon eegregnetregressiveemotionrecognitionincontinuousvadspaceusingeegsignals AT longbinjin eegregnetregressiveemotionrecognitionincontinuousvadspaceusingeegsignals AT hyuntaekjung eegregnetregressiveemotionrecognitionincontinuousvadspaceusingeegsignals AT hyunseokim eegregnetregressiveemotionrecognitionincontinuousvadspaceusingeegsignals AT eunyikim eegregnetregressiveemotionrecognitionincontinuousvadspaceusingeegsignals |