Research on Tri-Subspace Decoupling Clustering Graph Neural Network for EEG-Based Emotion Recognition

HUANG Chen; MA Hao-bo; ZHANG Yan; YANG Chao; SONG Jian-hua

doi:10.12263/DZXB.20250555

您当前的位置：

首页 >

文章列表页 >

Research on Tri-Subspace Decoupling Clustering Graph Neural Network for EEG-Based Emotion Recognition

PAPERS | 更新时间：2026-02-10

- Research on Tri-Subspace Decoupling Clustering Graph Neural Network for EEG-Based Emotion Recognition
- ACTA ELECTRONICA SINICA Vol. 53, Issue 11, Pages: 4065-4076(2025)
- 作者机构：
  
  1.湖北大学计算机学院，湖北武汉 430062
  2.智能感知系统与安全教育部重点实验室，湖北武汉 430062
  3.大数据智能分析与行业应用湖北省重点实验室，湖北武汉 430062
  4.湖北省高校人文社科重点研究基地-绩效评价信息管理研究中心，湖北武汉430062
  5.湖北大学网络空间安全学院，湖北武汉 430062
- 作者简介：
- 基金信息：
- DOI：10.12263/DZXB.20250555
  CLC： TP391;TP399
- Received：26 June 2025，
  
  Accepted：25 September 2025，
  
  Published：25 November 2025
- 稿件说明：
移动端阅览
黄辰, 马浩博, 张龑, 等. 用于脑电情绪识别的三子空间解耦聚类图神经网络研究[J]. 电子学报, 2025, 53(11): 4065-4076.

HUANG Chen, MA Hao-bo, ZHANG Yan, et al. Research on Tri-Subspace Decoupling Clustering Graph Neural Network for EEG-Based Emotion Recognition[J]. Acta Electronica Sinica, 2025, 53(11): 4065-4076.
黄辰, 马浩博, 张龑, 等. 用于脑电情绪识别的三子空间解耦聚类图神经网络研究[J]. 电子学报, 2025, 53(11): 4065-4076. DOI：10.12263/DZXB.20250555

HUANG Chen, MA Hao-bo, ZHANG Yan, et al. Research on Tri-Subspace Decoupling Clustering Graph Neural Network for EEG-Based Emotion Recognition[J]. Acta Electronica Sinica, 2025, 53(11): 4065-4076. DOI：10.12263/DZXB.20250555

摘要

图神经网络（Graph Neural Networks，GNNs）因其能够建模大脑区域间的时空依赖关系并捕捉上下文感知的神经模式，在基于脑电图（ElectroEncephaloGraphy，EEG）的情感识别中得到了广泛关注.然而，大多数基于GNN的EEG情感识别方法面临两个主要挑战：（1）许多现有模型未能考虑局部大脑区域间情感的共性和多样性，导致空间或功能相邻区域的节点嵌入过于同质化；（2）当前的方法通常依赖于简单的拼接或基于相关性的先验知识，这对于捕捉多个EEG通道和频带间复杂且分布的情感模式是不充分的.本文提出了一种三子空间解耦聚类图神经网络（Tri-Subspace-Decoupled Clustering Graph Neural Network，TS-DCGNN）来解决上述挑战.具体来说，TS-DCGNN将EEG信号解耦为三个子空间：显性情感子空间、隐性情感子空间和显性-隐性共振子空间，旨在捕捉可观察的情感体验（如“快乐”）、自动反应（如“惊吓”）及其耦合.此外，本文引入了一种双分支传播架构，其中图注意力网络（Graph Attention Networks，GATs）和图卷积网络（Graph Convolutional Networks，GCNs）并行操作，通过注意力驱动的交互和分层学习提取显性和隐性特征，从而增强区域情感表示.进一步地，本文提出了一个统一的表示学习模块，整合这些特征，并运用信息论方法来获得最小、充分和具有辨识度的情感表示.三个基准数据集上的实验表明，所提方法在性能上达到了最先进的水平，并提升了可解释性.

Abstract

Graph neural networks (GNNs) have gained significant attention in electroencephalography (EEG)-based emotion recognition for their ability to model spatial-temporal dependencies across brain regions and capture context-aware neural patterns. However

most GNN-based EEG emotion recognition methods encounter two primary challenges: (1) Many existing models fail to account for the emotional commonality and diversity across local brain regions

resulting in overly homogeneous node embeddings for spatially or functionally adjacent regions; (2) Current approaches often rely on simple concatenation or correlationbased priors

which are inadequate for capturing the complex and distributed emotional patterns across multiple EEG channels and frequency bands. In this paper

we propose a tri-subspace decoupling clustering graph neural network (TS-DCGNN) to address the above challenges. Specifically

TS-DCGNN decouples EEG signals into three subspaces: the explicit emotional

implicit emotional

and explicit-implicit resonance subspaces

aiming to capture observable experiences (e.g.

“happiness”)

automatic responses (e.g.

“startle”)

and their coupling. Moreover

we introduce a dual-branch propagation architecture where graph attention networks (GATs) and graph convolutional networks (GCNs) operate in parallel to extract explicit and implicit features via attention-driven interaction and hierarchical learning. This enhances regional emotional representations. Furthermore

we present a unified representation learning module that integrates these features and employs information theory to obtain a minimal

sufficient

and discriminative emotional representation. Experiments on three benchmark datasets demonstrate state-of-the-art performance and improved interpretability.

关键词

Keywords

references

HAEYEN S . A theoretical exploration of polyvagal theory in creative arts and psychomotor therapies for emotion regulation in stress and trauma [J ] . Frontiers in Psychology , 2024 , 15 : 1382007 .

AHAD A , JIANGBINA Z , TAHIR M , et al . 6G and intelligent healthcare: Taxonomy, technologies, open issues and future research directions [J ] . Internet of Things , 2024 , 25 : 101068 .

POURMIRZAEI M , ALI MONTAZER GA , MOUSAVI E . ATTENDEE: An affective tutoring system based on facial emotion recognition and head pose estimation to personalize e-learning environment [J ] . Journal of Computers in Education , 2025 , 12 ( 1 ): 65 - 92 .

MORIUCHI E , MURDY S . The role of robots in the service industry: Factors affecting human-robot interactions [J ] . International Journal of Hospitality Management , 2024 , 118 : 103682 .

AHUJA C , SETHIA D . TRANSIT-EEG: A framework for cross-subject classification with subject specific adaptation [J ] . IEEE Transactions on Cognitive and Developmental Systems , 2025 , 17 ( 4 ): 923 - 937 .

ZHANG D Z , CHEN F L , CHANG J L , et al . Structure aware multi-graph network for multi-modal emotion recognition in conversations [J ] . IEEE Transactions on Multimedia , 2024 , 26 : 3987 - 3997 .

ZHAO Y C , WANG G Y , SUN M X , et al . Methods for removing artifacts from EEG signals: A review [C ] // 2024 IEEE 24th International Conference on Software Quality, Reliability, and Security Companion . Piscataway : IEEE , 2024 : 752 - 758 .

JONKMAN A H , WARNAAR R S P , BACCINELLI W , et al . Analysis and applications of respiratory surface EMG: Report of a round table meeting [J ] . Critical Care , 2024 , 28 ( 1 ): 2 .

FAN X Y . Artifact removal from sleep-disordered EEG by wavelet enhanced independent component analysis [EB/OL ] . ( 2024-11 )[ 2025-06-24 ] . https://ualberta.scholaris.ca/server/api/core/bitstreams/8b0c4513-694e-4950-b200-7e5c69c0a6f8/content https://ualberta.scholaris.ca/server/api/core/bitstreams/8b0c4513-694e-4950-b200-7e5c69c0a6f8/content .

ZHANG G H , CARRASCO C D , WINSLER K , et al . Assessing the effectiveness of spatial PCA on SVM-based decoding of EEG data [J ] . NeuroImage , 2024 , 293 : 120625 .

WISNIEWSKI M G , JOYNER C N , ZAKRZEWSKI A C , et al . Finding tau rhythms in EEG: An independent component analysis approach [J ] . Human Brain Mapping , 2024 , 45 ( 2 ): e26572 .

LI Y R , LIU Y J , AI Y Q , et al . Explore the novel relationship of time-frequency domains in multivariate signals: A dynamic system perspective [J ] . IEEE Transactions on Bio-Medical Engineering , 2025 , 72 ( 6 ): 1872 - 1885 .

LALAWAT R S , BAJAJ V . An automatic framework for detecting autism spectrum disorder from EEG signals using TFD [J ] . IEEE Sensors Journal , 2024 , 24 ( 7 ): 10632 - 10639 .

GOMEZ-GIL F J , MARTÍNEZ-MARTÍNEZ V , RUIZ-GONZALEZ R , et al . Vibration-based monitoring of agro-industrial machinery using a k-Nearest Neighbors (kNN) classifier with a Harmony Search (HS) frequency selector algorithm [J ] . Computers and Electronics in Agriculture , 2024 , 217 : 108556 .

MAULANA B A , FAHMI M J , IMRAN A M , et al . Analisis sentimen terhadap aplikasi pluang menggunakan algoritma naive Bayes Dan support vector machine (SVM): Sentiment analysis of pluang applications with naive Bayes and support vector machine (SVM) algorithm [J ] . MALCOM: Indonesian Journal of Machine Learning and Computer Science , 2024 , 4 ( 2 ): 375 - 384 .

QI Y Q , JIA W C , FENG L L , et al . Piezoelectric touch sensing and random-forest-based technique for emotion recognition [J ] . IEEE Transactions on Computational Social Systems , 2024 , 11 ( 5 ): 6296 - 6307 .

BHAGAT D , VAKIL A , GUPTA R K , et al . Facial emotion recognition (FER) using convolutional neural network (CNN) [J ] . Procedia Computer Science , 2024 , 235 : 2079 - 2089 .

MANALU H V , RIFAI A P . Detection of human emotions through facial expressions using hybrid convolutional neural network-recurrent neural network algorithm [J ] . Intelligent Systems with Applications , 2024 , 21 : 200339 .

GUO Y Q , ZHANG B W , FAN X M , et al . A comprehensive interaction in multiscale multichannel EEG signals for emotion recognition [J ] . Mathematics , 2024 , 12 ( 8 ): 1180 .

LI C B , TANG T , PAN Y , et al . An efficient graph learning system for emotion recognition inspired by the cognitive prior graph of EEG brain network [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2025 , 36 ( 4 ): 7130 - 7144 .

KUANG W R , WANG Z , WEI Z W , et al . When transformer meets large graphs: An expressive and efficient two-view architecture [J ] . IEEE Transactions on Knowledge and Data Engineering , 2024 , 36 ( 10 ): 5440 - 5452 .

GAO Y Y , XUE Y F , GAO J . Emotion recognition from multichannel EEG signals based on low-rank subspace self-representation features [J ] . Biomedical Signal Processing and Control , 2025 , 99 : 106877 .

Saravia E . ML Visuals [EB/OL ] . [ 2025-06-24 ] . https://github.com/dair-ai/ml-visuals https://github.com/dair-ai/ml-visuals .

MARSICANO G , BERTINI C , RONCONI L . Decoding cognition in neurodevelopmental, psychiatric and neurological conditions with multivariate pattern analysis of EEG data [J ] . Neuroscience & Biobehavioral Reviews , 2024 , 164 : 105795 .

CHENG P Y , HAO W T , DAI S Y , J , et al . CLUB: A contrastive log-ratio upper bound of mutual information [C ] // Proceedings of the 37th International Conference on Machine Learning . Cambridge : PMLR , 2020 : 1779 - 1788 .

BELGHAZI M I , BARATIN A , RAJESHWAR S , et al . Mutual information neural estimation [C ] // Proceedings of the 35th International Conference on Machine Learning . Cambridge : PMLR , 2018 : 531 - 540 .

FEDERICI M , DUTTA A , FORRÉ P , et al . Learning robust representations via multi-view information bottleneck [EB/OL ] . ( 2020-02-18 )[ 2025-06-24 ] . https://arXiv.org/abs/2002.07017 https://arXiv.org/abs/2002.07017 .

SHAPIRO A . Monte Carlo sampling methods [M ] // Stochastic Programming . Amsterdam : Elsevier , 2003 : 353 - 425 .

FU Y H , OKADA S , WANG L B , et al . Context- and knowledge-aware graph convolutional network for multimodal emotion recognition [J ] . IEEE MultiMedia , 2022 , 29 ( 3 ): 91 - 100 .

LIU Y , GEDEON T , CALDWELL S , et al . Emotion recognition through observer’s physiological signals [EB/OL ] . ( 2020-02-19 )[ 2025-06-24 ] . https://arXiv.org/abs/2002.08034 https://arXiv.org/abs/2002.08034 .

KOELSTRA S , MUHL C , SOLEYMANI M , et al . DEAP: A database for emotion Analysis; Using physiological signals [J ] . IEEE Transactions on Affective Computing , 2012 , 3 ( 1 ): 18 - 31 .

ZHENG W L , LU B L . Investigating critical frequency bands and channels for EEG-based emotion recognition with deep neural networks [J ] . IEEE Transactions on Autonomous Mental Development , 2015 , 7 ( 3 ): 162 - 175 .

梁秀霞 , 夏曼曼 , 何月阳 , 等 . 基于时空多头图注意力网络的交通流预测 [J ] . 电子学报 , 2024 , 52 ( 2 ): 500 - 509 .

LIANG X X , XIA M M , HE Y Y , et al . Traffic flow prediction based on spatio-temporal multi-head graph attention network [J ] . Acta Electronica Sinica , 2024 , 52 ( 2 ): 500 - 509 . (in Chinese)

LAWHERN V J , SOLON A J , WAYTOWICH N R , et al . EEGNet: A compact convolutional neural network for EEG-based brain-computer interfaces [J ] . Journal of Neural Engineering , 2018 , 15 ( 5 ): 056013 .

HASIB M M , NAYAK T , HUANG Y F . A hierarchical LSTM model with attention for modeling EEG non-stationarity for human decision prediction [C ] // 2018 IEEE EMBS International Conference on Biomedical & Health Informatics . Piscataway : IEEE , 2018 : 104 - 107 .

SONG Y H , JIA X Y , YANG L , et al . Transformer-based spatial-temporal feature learning for EEG decoding [EB/OL ] . ( 2021-06-11 )[ 2025-06-24 ] . https://arXiv.org/abs/2106.11170 https://arXiv.org/abs/2106.11170 .

ZENG Y Y , ZHANG R R , YANG L , et al . Cross-domain text sentiment classification method based on the CNN-BiLSTM-TE model [J ] . Journal of Information Processing Systems , 2021 , 17 ( 4 ): 818 - 833 .

ZHONG P X , WANG D , MIAO C Y . EEG-based emotion recognition using regularized graph neural networks [J ] . IEEE Transactions on Affective Computing , 2022 , 13 ( 3 ): 1290 - 1301 .

LI X , LI J , ZHANG Y Z , et al . Emotion recognition from multi-channel EEG data through a dual-pipeline graph attention network [C ] // 2021 IEEE International Conference on Bioinformatics and Biomedicine . Piscataway : IEEE , 2021 : 3642 - 3647 .

PIKOULIS I , FILNTISIS P P , MARAGOS P . Leveraging semantic scene characteristics and multi-stream convolutional architectures in a contextual approach for video-based visual emotion recognition in the wild [C ] // 2021 16th IEEE International Conference on Automatic Face and Gesture Recognition . Piscataway : IEEE , 2021 : 1 - 8 .

LI R F , CHEN H , FENG F X , et al . DualGCN: Exploring syntactic and semantic information for aspect-based sentiment analysis [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2024 , 35 ( 6 ): 7642 - 7656 .

ZHU T , LI L D , YANG J F , et al . Multimodal emotion classification with multi-level semantic reasoning network [J ] . IEEE Transactions on Multimedia , 2023 , 25 : 6868 - 6880 .

LI G Q , CHEN N , JIN J . Semi-supervised EEG emotion recognition model based on enhanced graph fusion and GCN [J ] . Journal of Neural Engineering , 2022 , 19 ( 2 ): 026039 .

CHENG C , YU Z K , ZHANG Y , et al . Hybrid network using dynamic graph convolution and temporal self-attention for EEG-based emotion recognition [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2024 , 35 ( 12 ): 18565 - 18575 .

FU R R , CAI M P , WANG S W , et al . Exploring non-Euclidean relationships in EEG emotion recognition: A dynamic graph attention network [J ] . Biomedical Signal Processing and Control , 2024 , 94 : 106276 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

BottleneckNet: a Graph Neural Network for Post-Synthesis Timing Bottleneck Prediction in Large-Scale Digital ICs

STUDY OF THE ORIGIN OF INFRARED EMISSION IN GaAlAs DH RED LED's

RAMAN SCATTERING IN GaN GROWN BY VPE

USING Xe^1v-ION-LASER-INDUCED FLUORESCENCE TO STUDY THE PERFORMANCE OF ZnCdS:Ag FLUORESCENT SCREEN FOR MEDICAL X-RAY

THE (B(λkq) INTENSITY PARAMETERS AND 4f-4f TRANSITION PROBABILITIES OF Na₅Eu(MO₄)₄(M=Mo, W)

Related Author

HE Guang-peng

DI Zhi-xiong

DENG Yu-jiao

CHEN Xuan

ZHANG Ze-tao

LIU Yang

Related Institution

School of Information Science and Technology, Southwest Jiaotong University

School of Integrated Circuits, Southwest Jiaotong University

Faculty of Integrated Circuits, Xidian University

Hisilicon, Huawei Technologies Co., Ltd.

⁰