A robust adversarial learning framework for EEG emotion recognition

Abstract

Introduction: Electroencephalography (EEG) provides an objective means of capturing neural patterns associated with emotion, but substantial inter-subject variability limits the ability of existing models to generalize to unseen subjects.

Methods: We propose a robust adversarial domain generalization framework for cross-subject EEG emotion recognition. In the first stage, long short-term memory-based encoders are jointly trained to learn domain-invariant and domain-specific representations using reconstruction, emotion classification, and adversarial domain objectives. This process preserves emotion-relevant information while suppressing subject-specific distribution shifts. In the second stage, the invariant encoder and emotion classifier are further refined through distributionally robust optimization with gradient-guided feature augmentation. Theframework is evaluated on the AMIGOS and DREAMER datasets using leave-one-subject-out cross-validation.

Results: The proposed method achieves mean accuracies of 69.94% and 68.97% and F1 scores of 0.59 and 0.65 on AMIGOS and DREAMER, respectively, outperforming the comparison methods on both datasets. Ablation results further show that adversarial learning provides the primary improvement in cross-subject generalization, while distributionally robust optimization offers complementary robustness.

Discussion: These findings indicate that combining representation learning that is invariant across domains with distributionally robust optimization effectively reduces distribution shifts across subjects and improves EEG emotion recognition for unseen subjects.

Bilingual emotion recognition of social media texts using deep learning approach
Melkam Enyew Gashaw, Abebaw Alem, Anduamlak Abebe Fenta
Scientific Reports1 Jul 2026Peer-reviewedCC BYFull textGeneral medicineDOAJ
Explainable machine learning framework for dynamic line rating forecasting with adversarial robustness evaluation
Feyyaz Alpsalaz, Hasan Uzel, Yıldırım Özüpak +2
Scientific Reports1 Sep 2026Peer-reviewedCC BYFull textGeneral medicineDOAJ
Emotion Recognition Impairments in People with Dementia are Associated with Lower Physiological Linkage with Caregivers
Matthew Kiely, Ahria Dominguez, Casey K. Brown +3
Emotion (Washington, D.C.)22 Oct 2026Peer-reviewedLicence not statedFull textPsychiatry & mental healthPMC
Unmasking adversarial attacks using a robust XAI-driven approach for secure medical image classification
Sudarshan Saha, Sohel Rana, Ahmed Wasif Reza +1
Scientific Reports1 Jul 2026Peer-reviewedCC BYFull textGeneral medicineDOAJ
Facial emotion recognition of sadness relates to superior parietal gray matter volume in mild cognitive impairment.
Rachana Mahadevan, Naomi Kristin Giesers, Thomas Liman +4
Brain structure & function17 Oct 2026Peer-reviewedCC BYFull textPsychiatry & mental healthEurope PMC

Cite

BibTeX RIS

BibTeX

@article{yang2026robust,
  title = {{A robust adversarial learning framework for EEG emotion recognition}},
  author = {Hao Yang and Li Chen and Jianyong Gao and Junqiang Zhang and Jianfeng Zhang},
  journal = {Frontiers in neurology},
  year = {2026},
  month = oct,
  doi = {10.3389/fneur.2026.1951986},
  url = {https://europepmc.org/article/PMC/PMC13690544},
  note = {CC BY},
}

RIS

TY  - JOUR
TI  - A robust adversarial learning framework for EEG emotion recognition
AU  - Hao Yang
AU  - Li Chen
AU  - Jianyong Gao
AU  - Junqiang Zhang
AU  - Jianfeng Zhang
PY  - 2026
DA  - 2026/10/09
JO  - Frontiers in neurology
DO  - 10.3389/fneur.2026.1951986
UR  - https://europepmc.org/article/PMC/PMC13690544
AB  - Introduction: Electroencephalography (EEG) provides an objective means of capturing neural patterns associated with emotion, but substantial inter-subject variability limits the ability of existing models to generalize to unseen subjects.  Methods: We propose a robust adversarial domain generalization framework for cross-subject EEG emotion recognition. In the first stage, long short-term memory-based encoders are jointly trained to learn domain-invariant and domain-specific representations using reconstruction, emotion classification, and adversarial domain objectives. This process preserves emotion-relevant information while suppressing subject-specific distribution shifts. In the second stage, the invariant encoder and emotion classifier are further refined through distributionally robust optimization with gradient-guided feature augmentation. Theframework is evaluated on the AMIGOS and DREAMER datasets using leave-one-subject-out cross-validation.  Results: The proposed method achieves mean accuracies of 69.94% and 68.97% and F1 scores of 0.59 and 0.65 on AMIGOS and DREAMER, respectively, outperforming the comparison methods on both datasets. Ablation results further show that adversarial learning provides the primary improvement in cross-subject generalization, while distributionally robust optimization offers complementary robustness.  Discussion: These findings indicate that combining representation learning that is invariant across domains with distributionally robust optimization effectively reduces distribution shifts across subjects and improves EEG emotion recognition for unseen subjects.
SN  - 1664-2295
ER  -

Abstract

Related

Cite