Authors:
François Courtemanche
1
;
Aude Dufresne
1
;
Elise L. LeMoyne
2
and
Esma Aimeur
1
Affiliations:
1
University of Montréal, Canada
;
2
Tech3Lab, Canada
Keyword(s):
Affective Signal Processing, Temporal Construction, Psychophysiological Inference, Triangulation.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Biosignal Acquisition, Analysis and Processing
;
Data Manipulation
;
Health Engineering and Technology Applications
;
Human-Computer Interaction
;
Methodologies and Methods
;
Neurocomputing
;
Neurotechnology, Electronics and Informatics
;
Pattern Recognition
;
Physiological Computing Systems
;
Sensor Networks
;
Soft Computing
Abstract:
Predicting the psychological state of the user using physiological measures is one of the main objectives of physiological computing. While numerous works have addressed this task with great success, a large number of challenges remain to be solved in order to develop recognition approaches that can precisely and reliably feed human-computer interaction systems. This paper focuses on one of these challenges which is the temporal asynchrony between different physiological signals within one recognition model. The paper proposes a flexible and suitable method for feature extraction based on empirical optimisation of windows’ latency and duration. The approach is described within the theoretical framework of the psychophysiological inference and its common implementation using machine learning. The method has been experimentally validated (46 subjects) and results are presented. Empirically optimised values for the extraction windows are provided.