Deep Invertible Networks for EEG-based brain-signal decoding

TL;DR

This paper explores deep invertible networks for decoding EEG brain signals, demonstrating their ability to generate realistic signals and classify new data effectively, with discussions on improving their accuracy through regularization.

Contribution

It introduces the application of deep invertible networks to EEG decoding and discusses methods to enhance their performance.

Findings

Generated realistic EEG signals.

Achieved above-chance classification accuracy.

Discussed regularization techniques for better decoding.

Abstract

In this manuscript, we investigate deep invertible networks for EEG-based brain signal decoding and find them to generate realistic EEG signals as well as classify novel signals above chance. Further ideas for their regularization towards better decoding accuracies are discussed.