Back to publications

IEEE Open Journal of the Computer Society · 2026

EMO-CARE: EEG Multi-Scale Temporal Modeling with Channel-Aware Feature Attention for Robust Subject-Independent Emotion Recognition

View resource →

  • EEG
  • emotion recognition
  • deep learning
  • attention

Summary

EMO-CARE tackles one of the hardest problems in EEG-based emotion recognition: building models that generalize to new, unseen subjects rather than overfitting to the people they were trained on.

The framework combines multi-scale temporal modeling of EEG signals — capturing patterns at different time resolutions — with a channel-aware feature attention mechanism that learns which EEG electrodes carry the most emotion-relevant information. Together these produce more discriminative, robust representations for subject-independent emotion classification, which is what real clinical and consumer deployments actually require.

Authors

Y. Abdollahinejad, A. Mousavi, P. Siaplaouras, Z. Boukouvalas, R. Corizzo

Venue

IEEE Open Journal of the Computer Society, 2026.

Links