Home // SENSORCOMM 2015, The Ninth International Conference on Sensor Technologies and Applications // View article

Feature Selection and Interpretation of GSR and ECG Sensor Data in Biofeedback Stress Monitoring

Authors:
Tom Gedeon
Xuanying Zhu
Leana Copeland
Nandita Sharma

Keywords: observer stress; physiological signals; biofeedback; human centred computing

Abstract:
We have used sensor data previously to detect the stress status of observers of stressful environments. In this work we describe the process of conversion from a passive and post-facto detection of this stress status to a dynamic and real-time or close to real-time process for control of stress using biofeedback. We describe the changes required to feature selection and interpretation of galvanic skin response (GSR) and electrocardiograph (ECG) sensor data for the new setting, using training on data from the post- facto dataset filtered to match, and as far as possible simulate a real-time dataset. We then compare 3 alternatives including a control component.

Pages: 105 to 110

Copyright: Copyright (c) IARIA, 2015

Publication date: August 23, 2015

Published in: conference

ISSN: 2308-4405

ISBN: 978-1-61208-425-1

Location: Venice, Italy

Dates: from August 23, 2015 to August 28, 2015