Robust EEG-based Emotion Recognition Using an Inception and Two-sided Perturbation Model
Shadi Sartipi, Mujdat Cetin
TL;DR
Problem: DL-based EEG emotion recognition is vulnerable to environmental noise and adversarial perturbations. Approach: INC combines an Inception-based feature generator with two-sided perturbation (TSP) in an adversarial training framework to achieve subject-independent three-class emotion recognition. Key findings: robust accuracy up to $0.91 \pm 0.04$ under $L_2$ with PGD-10 and strong stability across perturbations, with superior performance compared to prior work; ablations show the effectiveness of TSP and AT. Impact: enhances reliability of EEG-based BCIs in noisy and adversarial environments, enabling safer real-time emotion decoding.
Abstract
Automated emotion recognition using electroencephalogram (EEG) signals has gained substantial attention. Although deep learning approaches exhibit strong performance, they often suffer from vulnerabilities to various perturbations, like environmental noise and adversarial attacks. In this paper, we propose an Inception feature generator and two-sided perturbation (INC-TSP) approach to enhance emotion recognition in brain-computer interfaces. INC-TSP integrates the Inception module for EEG data analysis and employs two-sided perturbation (TSP) as a defensive mechanism against input perturbations. TSP introduces worst-case perturbations to the model's weights and inputs, reinforcing the model's elasticity against adversarial attacks. The proposed approach addresses the challenge of maintaining accurate emotion recognition in the presence of input uncertainties. We validate INC-TSP in a subject-independent three-class emotion recognition scenario, demonstrating robust performance.