Graph-Embedded Convolutional Neural Network for Image-Based EEG Emotion Recognition

Emotion recognition from electroencephalograph (EEG) signals has long been essential for affective computing. In this article, we evaluate EEG emotion recognition by converting EEG signals from multiple channels into images such that richer spatial information can be considered and the question of E...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on emerging topics in computing Ročník 10; číslo 3; s. 1399 - 1413
Hlavní autoři: Song, Tengfei, Zheng, Wenming, Liu, Suyuan, Zong, Yuan, Cui, Zhen, Li, Yang
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York IEEE 01.07.2022
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Affective computing
Artificial neural networks
Brain modeling
Data mining
EEG emotion recognition
EEG image generation
Electrodes
Electroencephalography
Emotion recognition
Emotions
Feature extraction
graph convolutional neural network
graph embedded convolutional neural network
Image processing
Image recognition
Neural networks
Normal distribution
Object recognition
Performance evaluation
Spatial data
ISSN:2168-6750, 2168-6750
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Emotion recognition from electroencephalograph (EEG) signals has long been essential for affective computing. In this article, we evaluate EEG emotion recognition by converting EEG signals from multiple channels into images such that richer spatial information can be considered and the question of EEG-based emotion recognition can be converted into image recognition. To this end, we propose a novel method to generate continuous images from discrete EEG signals by introducing offset variables following a Gaussian distribution for each EEG channel to alleviate the biased electrode coordinates during image generation. In addition, a novel graph-embedded convolutional neural network (GECNN) method is proposed to combine the local convolutional neural network (CNN) features with global functional features to provide complementary emotion information. In GECNN, the attention mechanism is applied to extract more discriminative local features. Simultaneously, dynamical graph filtering explores the intrinsic relationships between different EEG regions. The local and global functional features are finally fused for emotion recognition. Extensive experiments in subject-dependent and subject-independent protocols are conducted to evaluate the performance of the proposed GECNN model on four datasets, i.e., SEED, SDEA, DREAMER, and MPED.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2168-6750
2168-6750
DOI:10.1109/TETC.2021.3087174