Authors:
Simon Ollander
;
Christelle Godin
;
Sylvie Charbonnier
and
Aurélie Campagne
Affiliation:
Univ. Grenoble Alpes, France
Keyword(s):
Stress, Features, Classification, Feature Selection, Sensor Selection, Driver Stress, Naive Bayes.
Related
Ontology
Subjects/Areas/Topics:
Affective Computing
;
Artificial Intelligence
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Biosignal Acquisition, Analysis and Processing
;
Data Manipulation
;
Health Engineering and Technology Applications
;
Health Information Systems
;
Human-Computer Interaction
;
Methodologies and Methods
;
Neurocomputing
;
Neurotechnology, Electronics and Informatics
;
Pattern Recognition
;
Physiological Computing Systems
;
Processing of Multimodal Input
;
Sensor Networks
;
Soft Computing
Abstract:
This study presents a real-life application-based feature and sensor relevance analysis for detecting stress in drivers. Using the MIT Database for Stress Recognition in Automobile Drivers, the relevance of various physiological sensor signals and features for distinguishing the driver’s state have been analyzed. Features related to heart rate, skin conductivity, electromuscular activity, and respiration have been compared using filter and wrapper selection methods. For distinguishing rest from activity, relevant sensors have been found to be heart rate, skin conductivity, and respiration (giving up to 94.6 ± 1.9 % accuracy). For distinguishing low stress from high stress, relevant sensors have been found to be heart rate and respiration (giving up to 78.1±4.1 % accuracy). In both cases, a multi-user model that requires only a calibration from the user in rest, without prior knowledge of the user’s individual stress dynamics, resulted in a different optimal sensor and feature configu
ration, giving 87.3±2.8 % and 72.1±4.3 % accuracy respectively.
(More)