Enhancing cross-subject EEG emotion recognition through multi-source manifold metric transfer learning

• Published in: Computers in biology and medicine Vol. 174; p. 108445
• Main Authors: Shi, XinSheng, She, Qingshan, Fang, Feng, Meng, Ming, Tan, Tongcai, Zhang, Yingchun
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.05.2024; Elsevier Limited
• Subjects: Accuracy; Adaptation; Affective brain-computer interface (aBCI); Affective computing; Algorithms; Brain - physiology; Brain-Computer Interfaces; Classification; Deep learning; EEG; Effectiveness; Electroencephalogram (EEG); Electroencephalography; Electroencephalography - methods; Emotion recognition; Emotions; Emotions - physiology; Euclidean geometry; Female; Fourier transforms; Human-computer interface; Humans; Internal Medicine; Machine Learning; Male; Manifolds (mathematics); Mapping; Methods; Metric transfer learning; Neural networks; Other; Signal Processing, Computer-Assisted; Statistical analysis; Support vector machines; Transfer learning; Wavelet transforms; Metric transfer learning; Emotion recognition; Electroencephalogram (EEG); Affective brain-computer interface (aBCI)
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2024.108445

Transfer learning (TL) has demonstrated its efficacy in addressing the cross-subject domain adaptation challenges in affective brain-computer interfaces (aBCI). However, previous TL methods usually use a stationary distance, such as Euclidean distance, to quantify the distribution dissimilarity between two domains, overlooking the inherent links among similar samples, potentially leading to suboptimal feature mapping. In this study, we introduced a novel algorithm called multi-source manifold metric transfer learning (MSMMTL) to enhance the efficacy of conventional TL. Specifically, we first selected the source domain based on Mahalanobis distance to enhance the quality of the source domains and then used manifold feature mapping approach to map the source and target domains on the Grassmann manifold to mitigate data drift between domains. In this newly established shared space, we optimized the Mahalanobis metric by maximizing the inter-class distances while minimizing the intra-class distances in the target domain. Recognizing that significant distribution discrepancies might persist across different domains even on the manifold, to ensure similar distributions between the source and target domains, we further imposed constraints on both domains under the Mahalanobis metric. This approach aims to reduce distributional disparities and enhance the electroencephalogram (EEG) emotion recognition performance. In cross-subject experiments, the MSMMTL model exhibits average classification accuracies of 88.83 % and 65.04 % for SEED and DEAP, respectively, underscoring the superiority of our proposed MSMMTL over other state-of-the-art methods. MSMMTL can effectively solve the problem of individual differences in EEG-based affective computing. •MSMMTL uses Mahalanobis distance to assess correlations between subjects and effectively screen suitable source domains.•MSMMTL utilizes the supervised information from the source domain to learn a more generalized distance representation.•This method constrains the distribution between domains on the Grassmann manifold and refines the feature mapping matrix.