Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions

Jonathan Shi Khai Ooi, Siti Anom Ahmad, Yu Zheng Chong, Sawal Hamid Md Ali, Guangyi Ai, Hiroaki Wagatsuma

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

An extensive variety of wellbeing frameworks had been introduced in modern vehicles a decade ago. Traction control, auto-braking, and anti-sleep systems are significant innovations that are presumed to be superior over human reaction. However, accident rates in Malaysia have yet to be fully reduced. In fact, in 2013, nearly one million enlisted vehicles were involved in road accidents, with damages reaching over RM9.3 billion. Meanwhile, a car is a system that encompasses the road, the vehicle, and the driver. At present, roads and vehicles have gained immense stability, but the driver remains as the most fragile component of this system. Electrodermal activity (EDA) was used in this study to investigate stress and anger as primary emotions leading to possible accidents involving the driver. A simulated driving assignment with preset neutral, stress, and anger scenarios was developed for emotional stimulation. A total of 20 subjects were included in this experiment. Acquired EDA signals were bandpass-filtered at 0.5 Hz to 2 Hz and subjected to short-Time Fourier transform. Then, their mean, median, and variance of power spectral density were extracted. The parameters obtained were statistically analyzed with two-sample f-Test. EDA readings from drivers demonstrated significant differences among neutral-stress, neutral-Anger, and stress-Anger simulated driving scenarios. The dataset was also divided into two groups (10-10) for training and testing of support vector machine classifier at 10-fold cross-validation. The classification accuracy was 85% each for neutral-stress and neutral-Anger and 70% for stress-Anger.

Original languageEnglish
Title of host publicationIECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages365-369
Number of pages5
ISBN (Electronic)9781467377911
DOIs
Publication statusPublished - 3 Feb 2017
Externally publishedYes
Event2016 IEEE-EMBS Conference on Biomedical Engineering and Sciences, IECBES 2016 - Kuala Lumpur, Malaysia
Duration: 4 Dec 20168 Dec 2016

Other

Other2016 IEEE-EMBS Conference on Biomedical Engineering and Sciences, IECBES 2016
CountryMalaysia
CityKuala Lumpur
Period4/12/168/12/16

Fingerprint

emotions
vehicles
accidents
roads
human reactions
Accidents
Traction control
Malaysia
sleep
braking
Highway accidents
traction
Power spectral density
Braking
classifiers
stimulation
Support vector machines
Fourier transforms
Classifiers
education

Keywords

  • anger
  • driving
  • electrodermal activity (EDA)
  • emotion
  • galvanic skin response (GSR)
  • stress

ASJC Scopus subject areas

  • Biomedical Engineering
  • Instrumentation

Cite this

Ooi, J. S. K., Ahmad, S. A., Chong, Y. Z., Md Ali, S. H., Ai, G., & Wagatsuma, H. (2017). Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions. In IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences (pp. 365-369). [7843475] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IECBES.2016.7843475

Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions. / Ooi, Jonathan Shi Khai; Ahmad, Siti Anom; Chong, Yu Zheng; Md Ali, Sawal Hamid; Ai, Guangyi; Wagatsuma, Hiroaki.

IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences. Institute of Electrical and Electronics Engineers Inc., 2017. p. 365-369 7843475.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ooi, JSK, Ahmad, SA, Chong, YZ, Md Ali, SH, Ai, G & Wagatsuma, H 2017, Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions. in IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences., 7843475, Institute of Electrical and Electronics Engineers Inc., pp. 365-369, 2016 IEEE-EMBS Conference on Biomedical Engineering and Sciences, IECBES 2016, Kuala Lumpur, Malaysia, 4/12/16. https://doi.org/10.1109/IECBES.2016.7843475
Ooi JSK, Ahmad SA, Chong YZ, Md Ali SH, Ai G, Wagatsuma H. Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions. In IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences. Institute of Electrical and Electronics Engineers Inc. 2017. p. 365-369. 7843475 https://doi.org/10.1109/IECBES.2016.7843475
Ooi, Jonathan Shi Khai ; Ahmad, Siti Anom ; Chong, Yu Zheng ; Md Ali, Sawal Hamid ; Ai, Guangyi ; Wagatsuma, Hiroaki. / Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions. IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 365-369
@inproceedings{7b54a88f821f4d7e90f67d5c7b696fff,
title = "Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions",
abstract = "An extensive variety of wellbeing frameworks had been introduced in modern vehicles a decade ago. Traction control, auto-braking, and anti-sleep systems are significant innovations that are presumed to be superior over human reaction. However, accident rates in Malaysia have yet to be fully reduced. In fact, in 2013, nearly one million enlisted vehicles were involved in road accidents, with damages reaching over RM9.3 billion. Meanwhile, a car is a system that encompasses the road, the vehicle, and the driver. At present, roads and vehicles have gained immense stability, but the driver remains as the most fragile component of this system. Electrodermal activity (EDA) was used in this study to investigate stress and anger as primary emotions leading to possible accidents involving the driver. A simulated driving assignment with preset neutral, stress, and anger scenarios was developed for emotional stimulation. A total of 20 subjects were included in this experiment. Acquired EDA signals were bandpass-filtered at 0.5 Hz to 2 Hz and subjected to short-Time Fourier transform. Then, their mean, median, and variance of power spectral density were extracted. The parameters obtained were statistically analyzed with two-sample f-Test. EDA readings from drivers demonstrated significant differences among neutral-stress, neutral-Anger, and stress-Anger simulated driving scenarios. The dataset was also divided into two groups (10-10) for training and testing of support vector machine classifier at 10-fold cross-validation. The classification accuracy was 85{\%} each for neutral-stress and neutral-Anger and 70{\%} for stress-Anger.",
keywords = "anger, driving, electrodermal activity (EDA), emotion, galvanic skin response (GSR), stress",
author = "Ooi, {Jonathan Shi Khai} and Ahmad, {Siti Anom} and Chong, {Yu Zheng} and {Md Ali}, {Sawal Hamid} and Guangyi Ai and Hiroaki Wagatsuma",
year = "2017",
month = "2",
day = "3",
doi = "10.1109/IECBES.2016.7843475",
language = "English",
pages = "365--369",
booktitle = "IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Driver emotion recognition framework based on electrodermal activity measurements during simulated driving conditions

AU - Ooi, Jonathan Shi Khai

AU - Ahmad, Siti Anom

AU - Chong, Yu Zheng

AU - Md Ali, Sawal Hamid

AU - Ai, Guangyi

AU - Wagatsuma, Hiroaki

PY - 2017/2/3

Y1 - 2017/2/3

N2 - An extensive variety of wellbeing frameworks had been introduced in modern vehicles a decade ago. Traction control, auto-braking, and anti-sleep systems are significant innovations that are presumed to be superior over human reaction. However, accident rates in Malaysia have yet to be fully reduced. In fact, in 2013, nearly one million enlisted vehicles were involved in road accidents, with damages reaching over RM9.3 billion. Meanwhile, a car is a system that encompasses the road, the vehicle, and the driver. At present, roads and vehicles have gained immense stability, but the driver remains as the most fragile component of this system. Electrodermal activity (EDA) was used in this study to investigate stress and anger as primary emotions leading to possible accidents involving the driver. A simulated driving assignment with preset neutral, stress, and anger scenarios was developed for emotional stimulation. A total of 20 subjects were included in this experiment. Acquired EDA signals were bandpass-filtered at 0.5 Hz to 2 Hz and subjected to short-Time Fourier transform. Then, their mean, median, and variance of power spectral density were extracted. The parameters obtained were statistically analyzed with two-sample f-Test. EDA readings from drivers demonstrated significant differences among neutral-stress, neutral-Anger, and stress-Anger simulated driving scenarios. The dataset was also divided into two groups (10-10) for training and testing of support vector machine classifier at 10-fold cross-validation. The classification accuracy was 85% each for neutral-stress and neutral-Anger and 70% for stress-Anger.

AB - An extensive variety of wellbeing frameworks had been introduced in modern vehicles a decade ago. Traction control, auto-braking, and anti-sleep systems are significant innovations that are presumed to be superior over human reaction. However, accident rates in Malaysia have yet to be fully reduced. In fact, in 2013, nearly one million enlisted vehicles were involved in road accidents, with damages reaching over RM9.3 billion. Meanwhile, a car is a system that encompasses the road, the vehicle, and the driver. At present, roads and vehicles have gained immense stability, but the driver remains as the most fragile component of this system. Electrodermal activity (EDA) was used in this study to investigate stress and anger as primary emotions leading to possible accidents involving the driver. A simulated driving assignment with preset neutral, stress, and anger scenarios was developed for emotional stimulation. A total of 20 subjects were included in this experiment. Acquired EDA signals were bandpass-filtered at 0.5 Hz to 2 Hz and subjected to short-Time Fourier transform. Then, their mean, median, and variance of power spectral density were extracted. The parameters obtained were statistically analyzed with two-sample f-Test. EDA readings from drivers demonstrated significant differences among neutral-stress, neutral-Anger, and stress-Anger simulated driving scenarios. The dataset was also divided into two groups (10-10) for training and testing of support vector machine classifier at 10-fold cross-validation. The classification accuracy was 85% each for neutral-stress and neutral-Anger and 70% for stress-Anger.

KW - anger

KW - driving

KW - electrodermal activity (EDA)

KW - emotion

KW - galvanic skin response (GSR)

KW - stress

UR - http://www.scopus.com/inward/record.url?scp=85015709163&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85015709163&partnerID=8YFLogxK

U2 - 10.1109/IECBES.2016.7843475

DO - 10.1109/IECBES.2016.7843475

M3 - Conference contribution

AN - SCOPUS:85015709163

SP - 365

EP - 369

BT - IECBES 2016 - IEEE-EMBS Conference on Biomedical Engineering and Sciences

PB - Institute of Electrical and Electronics Engineers Inc.

ER -