TY - GEN
T1 - Detection of Tactile Feedback on Touch-screen Devices using EEG Data
AU - Alsuradi, Haneen
AU - Pawar, Chaitali
AU - Park, Wanjoo
AU - Eid, Mohamad
N1 - Funding Information:
VI. ACKNOWLEDGEMENT This research is funded by the New York University Abu Dhabi PhD Fellowship Program.
Publisher Copyright:
© 2020 IEEE.
PY - 2020/3
Y1 - 2020/3
N2 - Neurohaptics strive to study brain activation associated with haptic interaction (tactile and/or kinesthetic). Understanding the haptic perception and cognition has become an exciting area in the technological, medical and psychophysical research. Neurohaptics has the potential to provide quantitative (objective) evaluation of the user haptic experience by directly measuring brain activities via EEG devices. In this study, we employed a Machine Learning (ML) based classifier model, namely the Radial Based Function Support Vector Machine (RBF-SVM) to select a few relevant Electroencephalography (EEG) channels and to detect the presence of tactile feedback during interaction with touch-screen devices using EEG data. To overcome the problem of limited training data, time-shifting is proposed as a method for data augmentation in time-series neural data which increased the classification accuracy. An experimental setup comprising an active touch task on the Tanvas touch-screen device is designed to evaluate the developed model. Results demonstrated that the middle frontal cortex, namely channels AF3, AF4, and F1 produced the best recognition rate of 85±3.3% in detecting the presence of the tactile feedback. This work is a step forward towards building a quantitative evaluation of tactile experience during haptic interaction.
AB - Neurohaptics strive to study brain activation associated with haptic interaction (tactile and/or kinesthetic). Understanding the haptic perception and cognition has become an exciting area in the technological, medical and psychophysical research. Neurohaptics has the potential to provide quantitative (objective) evaluation of the user haptic experience by directly measuring brain activities via EEG devices. In this study, we employed a Machine Learning (ML) based classifier model, namely the Radial Based Function Support Vector Machine (RBF-SVM) to select a few relevant Electroencephalography (EEG) channels and to detect the presence of tactile feedback during interaction with touch-screen devices using EEG data. To overcome the problem of limited training data, time-shifting is proposed as a method for data augmentation in time-series neural data which increased the classification accuracy. An experimental setup comprising an active touch task on the Tanvas touch-screen device is designed to evaluate the developed model. Results demonstrated that the middle frontal cortex, namely channels AF3, AF4, and F1 produced the best recognition rate of 85±3.3% in detecting the presence of the tactile feedback. This work is a step forward towards building a quantitative evaluation of tactile experience during haptic interaction.
UR - http://www.scopus.com/inward/record.url?scp=85085048019&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85085048019&partnerID=8YFLogxK
U2 - 10.1109/HAPTICS45997.2020.ras.HAP20.16.8d90d0bd
DO - 10.1109/HAPTICS45997.2020.ras.HAP20.16.8d90d0bd
M3 - Conference contribution
AN - SCOPUS:85085048019
T3 - IEEE Haptics Symposium, HAPTICS
SP - 775
EP - 780
BT - 2020 IEEE Haptics Symposium, HAPTICS 2020
PB - IEEE Computer Society
T2 - 26th IEEE Haptics Symposium, HAPTICS 2020
Y2 - 28 March 2020 through 31 March 2020
ER -