Deep Learning for Predicting Drivers Mental States from Electroencephalography (EEG)

Date

2017

Authors

Hajinoroozi, Mehdi

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The prediction and classification of the fatigue related mental states (drowsy/alert) of the drivers from electroencephalography (EEG) signals by means of deep learning methods and models is the concentration of this work. EEG data collected form drivers can be used for drivers' mental states (alert/drowsy states) prediction. The prediction of the drowsy and alert states of the drivers can be done by traditional machine learning methods but due to low performance of the traditional machine learning methods, it is desirable to examine new methods. Deep learning has already shown to be very successful in prediction of the complicated signals, especially in image processing field. In this work it is shown that different variation of deep learning models can be developed to effectively classify complicated EEG signals collected form drivers. In this work several novel variations of deep learning models have been developed and proposed for EGG data, related to drivers cognitive states, classifications which are able to considerably improve the predictions in comparison with traditional machine learning and custom deep learning models like convolutional neural networks. The developed and proposed deep learning models in this work can be eventually used in practical brain computer interface (BCI) systems which use EEG signals for brain and computer interactions.

Description

This item is available only to currently enrolled UTSA students, faculty or staff. To download, navigate to Log In in the top right-hand corner of this screen, then select Log in with my UTSA ID.

Keywords

Deep Learning, Deep neural networks, EEG classification, EEG prediction, Machine Learning

Citation

Department

Electrical and Computer Engineering