ST-M 2 EDA: spatio-temporal metric-guided multi-stream expert domain adaptation network for fatigue monitoring

In the human-machine collaboration paradigms of Industry 5.0, the intrinsic non-stationarity and domain drift of electroencephalogram (EEG) signals pose significant challenges to fatigue monitoring. Consequently, achieving accurate and robust cross-subject fatigue detection is paramount for workplace safety. This paper proposes a Spatio-Temporal Metric-Guided Multi-Stream Expert Domain Adaptation Network (ST-M 2 EDA). The framework initially employs Singular Value Decomposition (SVD)-based signal reconstruction to attenuate non-stationary noise, followed by Mixup data augmentation to enhance and smooth the feature space distribution. Furthermore, a hybrid architecture integrating Domain-Specific Batch Normalization (DSBN) with a ResNet-Transformer backbone is constructed to precisely decouple statistical discrepancies between domains. Finally, Multi-Kernel Maximum Mean Discrepancy (MK-MMD) is introduced as a dynamic routing criterion to facilitate fine-grained, adaptive decision-making for target samples. Experimental results on the SEED-VIG dataset demonstrate an average accuracy of 90.54%. Visual analysis of brain topography confirms that the model successfully aligns distributional shifts caused by individual heterogeneity-specifically correcting for specific frontotemporal lobe activation observed in subject S7-and establishes distinct decision boundaries. These findings indicate that synergizing signal reconstruction with Spatio-Temporal metric-guided multi-stream experts effectively mitigates negative transfer, ensuring reliable fatigue detection in complex cross-subject tasks.