TY - GEN
T1 - Image-based EEG signal processing for driving fatigue prediction
AU - Cheng, Eric Juwei
AU - Young, Ku Young
AU - Lin, Chin Teng
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/2
Y1 - 2018/7/2
N2 - This study proposes a EEG-based prediction system that transform the measured EEG record into an image-liked data for estimating the drowsiness level of drivers. Drowsy driving is one of the main factors to the occurrence of traffic accident. Since drivers themselves may not always immediately recognize that they are in the drowsy state, the risk of traffic accident increases while the driver is in the low vigilance state. In order to address this problem, the estimation of drowsy driving state via brain-computer interfaces (BCI) becomes a major concern in the driving safety field. This study transforms the measured EEG record into a image-liked feature maps, and then passes these feature maps to a Convolutional Neural Network (CNN) to learn the discriminative representations. The proposed drowsiness prediction system is evaluated by leave-one-subject-out cross-validation. The results indicate that our approach provides impressive and robust prediction performance on the EEG dataset without artifact removal process.
AB - This study proposes a EEG-based prediction system that transform the measured EEG record into an image-liked data for estimating the drowsiness level of drivers. Drowsy driving is one of the main factors to the occurrence of traffic accident. Since drivers themselves may not always immediately recognize that they are in the drowsy state, the risk of traffic accident increases while the driver is in the low vigilance state. In order to address this problem, the estimation of drowsy driving state via brain-computer interfaces (BCI) becomes a major concern in the driving safety field. This study transforms the measured EEG record into a image-liked feature maps, and then passes these feature maps to a Convolutional Neural Network (CNN) to learn the discriminative representations. The proposed drowsiness prediction system is evaluated by leave-one-subject-out cross-validation. The results indicate that our approach provides impressive and robust prediction performance on the EEG dataset without artifact removal process.
UR - http://www.scopus.com/inward/record.url?scp=85062388915&partnerID=8YFLogxK
U2 - 10.1109/CACS.2018.8606734
DO - 10.1109/CACS.2018.8606734
M3 - Conference contribution
AN - SCOPUS:85062388915
T3 - 2018 International Automatic Control Conference, CACS 2018
BT - 2018 International Automatic Control Conference, CACS 2018
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2018 International Automatic Control Conference, CACS 2018
Y2 - 4 November 2018 through 7 November 2018
ER -