# Adversarial Deep Learning in EEG Biometrics

## TL;DR

This paper introduces an adversarial deep learning approach to EEG biometrics that learns session-invariant representations, improving robustness and longitudinal usability for person identification.

## Contribution

It proposes a novel adversarial inference method within deep convolutional networks to enhance session-invariance in EEG-based person identification.

## Key findings

- Improved identification accuracy with longitudinal EEG data
- Robustness to session variability demonstrated
- Effective adversarial training for invariant feature learning

## Abstract

Deep learning methods for person identification based on electroencephalographic (EEG) brain activity encounters the problem of exploiting the temporally correlated structures or recording session specific variability within EEG. Furthermore, recent methods have mostly trained and evaluated based on single session EEG data. We address this problem from an invariant representation learning perspective. We propose an adversarial inference approach to extend such deep learning models to learn session-invariant person-discriminative representations that can provide robustness in terms of longitudinal usability. Using adversarial learning within a deep convolutional network, we empirically assess and show improvements with our approach based on longitudinally collected EEG data for person identification from half-second EEG epochs.