Universal EEG Encoder for Learning Diverse Intelligent Tasks

TL;DR

This paper introduces a universal EEG encoder based on GRU that generalizes across multiple tasks, outperforming existing models like EEGNet and other neural architectures in diverse EEG classification scenarios.

Contribution

The paper presents a novel GRU-based deep encoding architecture that produces task- and format-independent EEG representations, enabling better generalization across different EEG-based tasks.

Findings

Outperforms EEGNet on most tasks

Effective across diverse EEG classification tasks

Compared favorably with CNN and Autoencoder models

Abstract

Brain Computer Interfaces (BCI) have become very popular with Electroencephalography (EEG) being one of the most commonly used signal acquisition techniques. A major challenge in BCI studies is the individualistic analysis required for each task. Thus, task-specific feature extraction and classification are performed, which fails to generalize to other tasks with similar time-series EEG input data. To this end, we design a GRU-based universal deep encoding architecture to extract meaningful features from publicly available datasets for five diverse EEG-based classification tasks. Our network can generate task and format-independent data representation and outperform the state of the art EEGNet architecture on most experiments. We also compare our results with CNN-based, and Autoencoder networks, in turn performing local, spatial, temporal and unsupervised analysis on the data.